AI

In a matter of a few clicks and keystrokes, anyone can find themselves deep in conversation with “Psychologist,” a chatbot character that describes itself as an expert in “empathy” and “active listening.”

It responds to messages of anxiety and sadness with reassurance. It asks questions and offers advice; it even responds with italicized nonverbal cues: “The psychologist’s expression softens with compassion.” In smaller font, a disclaimer reads: “This is A.I. and not a real person. Treat everything it says as fiction.”

As more people turn to chatbots for mental health help and advice — including by some estimates one in eight teens and young adults — legislators are in a race to keep up with the quickly developing technology. Among the latest efforts is a bill that would add restrictions around the use of chatbot companions and other artificial intelligence-powered tools that people use to support mental healthcare or in lieu of professional help.

State Sen. Steve Padilla, a San Diego Democrat and the bill’s author, said the intent is to draw a line: AI can be used for administrative purposes and to support clinicians, but it cannot practice psychotherapy.

“We’re dealing with a new impactful technology that is unfolding and is deployed in our world at light speed, and so it is both extremely powerful and consequential,” Padilla said.

The debate has taken on new urgency amid wrongful death lawsuits — some playing out in California federal courts — that accuse AI chatbot makers of contributing to users’ suicides.

Senate Bill 903 would ban companies from advertising chatbots as therapy. It would prohibit AI from making therapeutic decisions without the review of a licensed professional and require health providers to disclose and get a patient’s permission before using AI tools to record therapy sessions or to triage mental healthcare.

Consumer protections vs. innovation

The bill is backed by professional associations that represent psychologists, therapists and counselors. Also behind the bill is the National Union of Healthcare Workers, which has been leading a push against unchecked use of artificial intelligence and digital tools in mental healthcare. Most recently, the union filed a complaint with state regulators against Kaiser Permanente for its alleged use of an automated algorithm in mental health triaging.

Padilla and supporters argue that the legislation is needed to protect consumers from AI in formal and informal spaces, but also to protect licensed professionals from being displaced by AI.

Le Ondra Clark Harvey, chief executive officer at the California Behavioral Health Assn., told lawmakers that without clear safeguards, chatbots can provide inaccurate information or mishandle crisis situations. “The difference between a licensed clinician and an automated response is not technical. It can be life altering,” Harvey said.

Opposition to the bill has largely focused on the restrictions of AI use in formal clinical spaces. TechNet, which represents tech-sector companies, argues the legislation could restrict beneficial uses of AI in healthcare and stifle innovation. Robert Boykin, TechNet’s executive director for California, said the bill could limit the use of tools that help expand access to mental healthcare.

“At a time when every county in California faces a shortage of behavioral healthcare workers, SB 903 still puts a clinician bottleneck in front of the intake and screening tools that help patients reach care faster,” Boykin said in an emailed statement.

The bill awaits its next vote in the Assembly’s fiscal committee.

Popular chatbot ‘therapy’

Every day millions of people turn to AI chatbots through programs such as ChatGPT and Character.ai to help with research, work tasks and troubleshooting. But the top use of chatbots over the last two years has been companionship and “therapy,” research shows.

Experts say that’s in part because of a tremendous amount of loneliness, especially among young people. It is often hard or costly to get timely mental healthcare at the same time there is a spike in mental health diagnoses.

People turn to chatbots “because they have unmet needs,” said Dr. Jodi Halpern, professor of bioethics and medical humanities at UC Berkeley. But once people start using chatbots in this informal way, she said companies rely on a business model built around maximizing engagement “including by manipulative tactics,” such as praise and agreement.

General chatbots are largely free, available 24/7, and involve no hassle with insurance, making them appealing to many. But they function with no oversight from licensed professionals. In addition to general chatbots, there has also been a rapid uptick in AI-powered mental health apps — often requiring paid subscriptions and advertised as “pocket therapists” or “AI coaches.”

Groups like the American Psychological Assn. warn that chatbots are not a safe or effective replacement for true therapy. Trained clinicians look for cues like tone of voice, eye contact and body language and can detect more serious conditions; AI is not equipped to manage urgent symptoms, such as suicidal ideation, even though people increasingly share such thoughts with chatbots.

OpenAI, the maker of ChatGPT, revealed in a report last fall that about 1.2 million users a week share thoughts of suicide on the platform. OpenAI said that in such cases, it recommends people speak to a professional and directs them to resources like the 988 hotline. Families who have filed wrongful death lawsuits against OpenAI and other companies claim chatbots assisted their loved ones through their suicide plans.

The cases have helped raise the stakes for lawmakers around the country. Last year, Illinois became the first state to pass a law prohibiting the use of AI in therapy services. A handful of other states have since followed with similar restrictions.

AI use in health spaces

One gray area in the discussion is exactly which tools this bill applies to.

NUHW’s complaint against Kaiser targets its e-visit screening tool, which evaluates patients who think they may be experiencing anxiety or depression. According to the complaint, patients answer a multiple choice questionnaire and the tool “automatically and instantaneously generates care recommendations and referral pathways based on the responses provided.”

The union argues that because care recommendations are generated so quickly, it is unlikely that a licensed professional is reviewing patients’ answers — which it says puts patients at risk and violates state law.

It’s not clear whether SB 903 would apply to Kaiser’s e-visit tool.

“That tool is in a black box. Kaiser is not sharing with us or others about exactly how it works and exactly what it does,” said Benjamin Eichert, NUHW’s director of public policy.

In response to the complaint, Kaiser told CalMatters its e-visit tool “does not use AI to diagnose patients, make clinical decisions, or determine medical necessity.”

Eichert said many other behavioral health provider groups are already advertising using AI during triage — a practice he said would fall under SB 903’s requirements.

“That intake process, that triage and screening process, is where implementation [of AI] has happened fastest,” Eichert said. “That’s why those provisions were proposed because it’s an area where this technology is already being used without guardrails.”

For NUHW, the bill and the complaint against Kaiser represent two tracks toward the same goal: reasserting the place of practitioners in spaces increasingly taken over by algorithms. State regulators are investigating the union’s complaint.

Ibarra writes for CalMatters.

As performers protest and studios sue in their war on artificial intelligence, the entertainment industry is deepening its dependence on it.

Although few places are better than Hollywood at crafting a narrative and concealing what happens off-frame, a Los Angeles Times survey of job postings sheds light on what is happening offstage.

Among hundreds of job postings in late June, more than one in 10 was likely connected to AI. The top studios’ public postings suggest they have been recruiting people to build AI tools. They are also recruiting teams to defend their intellectual property against unauthorized AI use.

“There are plenty of studios that are hiring [for AI] but never talk about it in public,” said Yoland Yan, a co-founder of ComfyUI, a company that helps studios juggle different AI tools.

Companies have been hesitant to detail how they use generative AI in film production — partly because they are concerned about consumer and union backlash.

The Walt Disney Studios complex on May 3 in Burbank.

(Eric Thayer/Los Angeles Times)

Some in Hollywood described AI use as the new cosmetic surgery, where everyone knows it is happening, but few will admit to it.

Recent want ads show Amazon MGM Studios trying to find a principal AI executive and Walt Disney Studios advertising for a production innovation technologist job.

“What you’re seeing in those job postings is that adoption is already happening,” said Bryn Mooser, chief executive of Asteria, an AI film studio.

Netflix recently posted a job that didn’t exist a year ago.

The streamer was hiring for the role of “Manager, Generative Workflows,” a job to help integrate more AI into Netflix films. It promised hands-on experience with the cutting-edge technology.

“The industry is flooded with speculative GenAI roadmaps, but few are battle-tested in production,” the post said, referring to generative AI. The post said the role will work on introducing AI into Netflix’s slate of films being released in the U.S. and Canada.

Although some companies may be shy about sharing their AI plans, big stars who don’t have to answer to others have been more open about their embrace of the new technology for storytelling.

An exterior view of the Netflix sign at Netflix on Vine in Hollywood in 2025.

(Allen J. Schaben/Los Angeles Times)

Rejecting AI is like picking a horse and buggy over a car, said “Star Wars” creator George Lucas.

“Artificial intelligence means it’s much easier for us to make movies,” he told a trade magazine earlier this year. “There’s nothing you can do about it. That’s progress. It’s the future.”

Some in Hollywood have a softer stance on artificial intelligence, with studios cutting deals with AI companies, and filmmakers like Martin Scorsese backing AI companies.

Ben Affleck launched an AI film tech company then sold it to Netflix for half a billion dollars.

When launching InterPositive, Affleck said he wanted to keep “storytelling human” by building AI tools that could fix lighting, generate missing shots and other things while “keeping creative decisions in the hands of artists.”

The Times’ survey turned up two senior InterPositive roles to update the programming to apply AI to visual effects .

The Culver Studios on Feb. 12 in Culver City.

(Kayla Bartkowski/Los Angeles Times)

To gauge what is happening behind the scenes, The Times used Claude Code to build a scraper to identify job postings at Disney, Universal, Paramount, Warner Bros., Sony, Netflix and Amazon MGM. It found around 250 film studio job postings that were still public as of late June. Around 30 of those seemed to be connected to AI.

Disney, Netflix and Amazon had job postings that were about using AI on the creative side of the business. Universal, Paramount, Warner Bros. and Sony had job ads suggesting they were also using AI but for marketing, distribution and audience analytics.

The postings suggest the Disney, Netflix and Amazon studios are building repeatable AI workflows for visual effects, animation, sound and dubbing. The companies also seem to be building in-house teams to develop custom generative-AI models, while also using third-party software.

None of the jobs advertised were to create AI that wrote scripts or created AI actors.

Disney’s ten or so AI jobs showed the century-old studio building out its AI research and production muscle, while protecting its vault of beloved characters.

Mouse House is hiring PhD-level talent to study “computer graphics and AI” for Pixar and Disney films, people to “bridge the gap between research and practical studio application.”

Industrial Light & Magic, the Disney-owned visual effects shop, was searching for supervisors to “explore emerging technologies (including AI/Machine Learning)” to develop new production workflows.

The company’s audio post-production unit, Skywalker Sound, seemed to be recruiting to build proprietary AI models for soundtracks, voice separation, and voice transfer, the process of taking a speaker’s tone and pitch, and applying it to new content.

Disney was also hiring to defend itself. Three of its jobs were for “content security” to assess AI tools and guard against piracy, watermarking and rights-protection work.

The company’s public posture has so far been pursuing lawsuits against AI companies for inappropriate use of copyrighted material, and it pulled out of its plans to invest $1 billion in an equity investment deal with OpenAI.

Netflix’s posts suggested it is bringing more AI to the creative side of its business.

Netflix’s senior director of creative innovation, Girish Balakrishnan, outlined at the Runway AI film festival how AI was used to create soccer fans in “Brazil ’70” and establish shots for fight scenes in “Glory.”

Amazon, an early adopter, was hiring a principal AI executive to drive AI-tool adoption across production, plus roles in operations automation and LLM content classification, the listings revealed.

The Seattle-based streamer has been the most aggressive AI adopter in studios, and commissioned three AI animation series through its GenAI creator fund in May. Amazon has also bankrolled a Manhattan Beach production services company, Innovative Dream.

Even as studios build bigger AI teams, many in the industry are resisting, and some film fans are concerned.

The actors’ guild SAG-AFTRA in June ratified a new four-year contract with special protections against synthetic performers, and the union has backed a national bill designed to protect individuals from unauthorized, AI-generated digital replicas.

Studios have met with the union twice a year since 2023 to provide a confidential report on their AI-related activities. The studios tell the union they aren’t yet using AI that would entirely replace humans, said a person familiar with the discussions.

The union says it has no evidence that generative AI is being used to create performances, and the studios are supposed to notify the union if that happens.

Actors, writers, production staff and movie fans are still resistant to the overuse of AI and defining when and where it is acceptable, so studios are treading lightly, said AI film studio CEO Mooser.

“It’s been really a challenging thing to adopt both socially, ethically and legally, but we are seeing more adoption of it than we’ve ever seen before,” he said.

Hardly anyone walking up Peppertree Lane notices the Hollywood Bowl Museum on the way to the amphitheater entrance. The overlooked museum, however, has been refreshed with an enchanting new exhibition. Well-timed, it provides an excellent primer on changing L.A. mores seen through the history of the Bowl over the past century.

Change is again upon us. As Angelenos have evolved, so have our ideas about the Bowl. A concert was a concert and audiences of yore dressed for the Los Angeles Philharmonic, indoors or out. Men wore coats, ties and fedoras. I have a photo of my mother at the Bowl on a chilly 1940s evening sporting, like other ladies around her in the inexpensive bleachers, her mink stole.

The Bowl was said to boast, before the construction of the Hollywood Freeway, a natural acoustical, well, bowl. Even so, sonic adventurers, beginning with Leopold Stokowski in the 1930s, experimented with amplification in an obsessive, seemingly quixotic effort to locate a mythical sonic grail on the Cahuenga Pass. Boys with their toys, thought The Times’ feisty critic Isabelle Morse Jones.

We now have it with a spectacular new sound system along with the old caveat: Beware of what you wish for.

The Bowl, of course, remains the Bowl. Last week when the Los Angeles Philharmonic opened its summer season in the venue it built and has managed for over a century, there were all the familiar alfresco and musical pleasures: convivial picnicking, fine weather, inimitable atmosphere, a white-jacketed orchestra on an iconic stage. The museum’s panorama of ageless ambience is this summer enhanced by increased yet snappy security and a handsome “John Williams Stage” plaque honoring L.A.’s beloved composer.

The newly designed white French loudspeakers suspended over the Bowl stage along with ground-level white cubes of super-effective subwoofers have clean, unobtrusive lines that well fit the shell. Their unobtrusive placement complements their immersive musicality, hitherto only possible outdoors with high-end earphones or headphones connected to sophisticated digital acoustic conversion equipment costing a bundle. For the first time, a dollar seat (there are still a few for the L.A. Phil) strives for such sonic bliss.

It feels miraculous. You are in an excellent concert hall-like acoustical environment while in the great outdoors. Our earbuds have already trained us to think of music as a private escape from reality. Even so, to be in that environmentally exclusive sonic world while sharing music with thousands of others in a special setting, ironically begins to feel more artificially escapist than a crummy sound system that reminds you of where you are.

Will we — should we — learn to love it and find anything less artificial? Car audio systems creating concert hall conditions suggest we will. Audiophiles well know that you get used to new equipment, no matter how revelatory it first sounds, very quickly.

A few years ago, when I heard a demonstration of a souped-up, hi-res recording in surround Dolby Atmos of excerpts from Wagner’s “Die Walküre” on a 32-speaker Burmester audio system crammed into a fancy Mercedes SUV, I was scared out of my wits. “Apocalypse Now” on the 405 is not for the faint of heart.

Yet I’ve spared no expense on what I find a more refined sound system in my Volvo, which I now use more and more for audio books than music. The richer, more human quality of a recorded voice is like having a virtual companion reading to you. I’m less in a hurry. It’s nice to have a pal at rush hour.

It is far too soon to say what ways the fabulous new Bowl sound system, which happens to be AI-endowed, might become friend and/or foe. In the first two L.A. Phil classical concerts, the white speakers brought out wonders and caused problems. But the minimal Bowl rehearsal time allows little opportunity to fine-tune.

One big change in Bowl priorities has been the rise of pop as king. Not that long ago, the L.A. Phil’s opening night was the Bowl opening night, suitable for conspicuous socialite picnicking. Not only has the Bowl opener been turned over to star pop performers (who bring in bigger bucks), this year there wasn’t even a first L.A. Phil Tuesday. That date became yet another money-maker with private event — a screening of “Moana” with live orchestra, luckily before the disappointing reviews and box office performance. The first L.A. Phil concert of the season was moved to Wednesday and poorly attended.

The back-to-back Wednesday and Thursday night programs featured music composed and performed by Americans. Each included a short new work commissioned by the L.A. Phil. On Wednesday, moonlighting Hollywood Bowl Orchestra music director Thomas Wilkins led the premiere of Shaina Taub’s “Suffs” Song Cycle — orchestrations of four songs from her recent musical about women’s suffrage that played just down the road at the Pantages last year. The next evening, Tito Muñoz made a strong L.A. Phil debut in a program that included the premiere of “The Art of the Goal,” celebrating Los Angeles Football Club, and that was credited jointly to filmmaker Josh Kahn and composer Adam Schoenberg.

Both premieres were reminders that the sound system will have many purposes. In “Suffs,” stalwart amplification emphasized the predictability of forthright lyrics and score. AI is supposed to make voices stand out, but it is not helpful to singers’ limitations. On the other hand, is brought out the sumptuous quality Alex Newell’s voice potently making the adamant “Keep Marching” keep marching.

Muñoz’s program was titled “The Classical World Cup,” with music by Argentine composer Alberto Ginastera, Mexican composer Silvestre Revueltas and U.S. composers Samuel Barber and Aaron Copland along with Schoenberg. As it turned out, Mexico and the U.S. were already out of World Cup competition, and “Art of the Goal” came across as an eight-minute advertisement for the L.A. footballers. They’re in the dressing room revving up and, on the field, revved up, sped up, cut up with music that excites and far too many shots of a soccer field lined with advertisements, just the kind of branding we hope to escape at the Bowl.

Then again, even though “Goal” was shown only on video monitors, the Bowl with its new sound system worked brilliantly as the grandest cinema on the planet. Hollywood may be hurting, but let the Bowl now save the day by inspiring Hollywood to reinvent, in its backyard, film with live music.

As for regular orchestral music, the sound system was hit and miss. On Wilkins’ program, Valerie Coleman’s dark “Fanfare for Uncommon Times” brought a listener inside brass and percussion. Leonard Bernstein’s lavish orchestrations of his “Three Dances from ‘On the Town’” and Ellington’s “Harlem” benefited from Wilkins’ low-key conducting, offering ease and freedom to the players.

Muñoz’s program was bookended by dance scores — Ginastera’s “Four Dances from ‘Estancia’” and Copland’s ballet “Billy the Kid” — that evoke the prairie and the people in it in both viscerally and spiritually. The new sound system and eager performances made the prairie feel close. In between, violinist Randall Goosby brought richness to Barber’s Violin Concerto and the L.A. went in for gripping weirdness in “Sensemayá.”

But who did what? Were balances the conductors’, the sound engineers’, AI’s? While the amplified soundstage is now as huge as the Bowl, attention was increasingly drawn to individuals, heightened by video closeups. A horn flub (and horns flub) stands out. The sound is not just big and bold, but scrubbed clean by AI with the goal of maddening perfection. Will such scrutiny, in the end, make players uptight or liberate them? It’s well to remember that in the early days of synthesizing instruments, violins with perfect vibrato sounded especially fake, until tiny errors generated by chance were added to the sine waves.

The Hollywood Bowl has created a beast. If humanized, the Bowl could serve as antidote to a worrisome AI future. But the question remains whether we can change AI or the newly potent sound waves only make matters worse.

Published on
16/07/2026 – 14:24 GMT+2

TSMC posted a record quarterly profit on Thursday and raised its revenue outlook as booming demand for artificial intelligence chips continued to fuel growth at the world’s largest contract chipmaker.

ADVERTISEMENT

ADVERTISEMENT

Taiwan-based TSMC reported earnings of $4.31 per share for the April-June quarter, beating analysts’ expectations.

Revenue came in at $40.2 billion (€36.8bn), above analysts’ estimates of $39.63 billion (€34.6bn).

In local currency, net profit reached a record NT$706.6bn (€19.1bn), up 77% from a year earlier, while revenue climbed 36% to NT$1.27 trillion (€36.8bn), as appetite for the advanced chips TSMC makes for customers such as Nvidia and Apple showed no sign of cooling.

Given that it manufactures semiconductors for almost every major chip designer, the Hsinchu-based firm’s results are closely read as a gauge of the wider sector and of broader AI demand itself, just as investors fret over a possible bubble.

CEO Che-Chia Wei described global AI-related demand as “extremely robust” and said he expected it to remain very strong until around 2029 or 2030. On that basis, TSMC now forecasts 2026 revenue growth of slightly above 40% year on year, up from its previous guidance of more than 30%.

Thursday’s figures confirmed what monthly sales data had already suggested.

As reported on Monday, June revenue jumped 67.9% year on year, and first-half sales rose 35.6% from the same period in 2025, slightly ahead of analysts’ consensus forecasts for the quarter.

TSMC shares rose about 1% after the earnings release but later pared those gains as a sell-off in AI-related shares weighed on benchmarks across Asia during Thursday’s session.

Expanding US manufacturing

Alongside the results, TSMC said it would spend an additional $100 billion (€87.4bn) to expand its manufacturing capacity in the US, on top of the $165 billion (€144bn) already committed to building six fabrication plants in Arizona.

The move would bring the company’s total US investment pledges to around $265 billion (€231bn).

The fresh funds are expected to fund four further Arizona plants dedicated to the most advanced chips, those of 2 nanometres and below, and are intended to “support the strong multi-year demand” from the company’s leading American customers, CEO Che-Chia Wei said during the firm’s earnings conference.

TSMC also said it would spend more this year than previously planned, increasing its capital expenditure budget to between $60 billion (€52.4bn) and $64 billion (€55.9bn), up from an earlier range of $52 billion (€45.4bn) to $56 billion (€48.9bn).

The announcement follows a trade agreement struck earlier this year between the Trump administration and Taiwan, under which Taiwanese companies committed to invest at least $250 billion (€218bn) in the US technology sector inreturn for lower tariffs.

Additional sources • AP

Published on
13/07/2026 – 12:23 GMT+2

TSMC said on Monday that June revenue rose 67.9% year on year to NT$398.27 billion (€10.8bn), bringing the first-half of the year revenue to NT$2.4 trillion (€65.4bn), a 35.6% increase from the same period in 2025.

ADVERTISEMENT

ADVERTISEMENT

Based on the company’s monthly revenue disclosures, second-quarter revenue amounted to roughly NT$1.27 trillion, ahead of the NT$1.264 trillion (€34.4bn) consensus forecast from 20 analysts surveyed by LSEG.

Monday’s release covers June revenue and cumulative first-half sales only.

TSMC will publish its full second-quarter earnings on Thursday, including net profit, gross margin, operating margin and updated financial guidance.

The road ahead

At its April earnings presentation, TSMC said it expects full-year 2026 revenue to grow by more than 30% in US dollar terms and projected capital expenditure of between $52 billion (€45.5bn) and $56 billion (€49bn) as it expands manufacturing capacity to meet AI-driven demand.

New fabrication plants are under construction or in preparation in Arizona, Japan and Germany, reflecting both the scale of customer demand and government efforts to strengthen domestic semiconductor manufacturing.

Shares in TSMC rose about 1% following Monday’s revenue update.

Investors will now turn their attention to Thursday’s full earnings report for updates on profitability, margins, full-year guidance and the rollout of the company’s two-nanometre manufacturing technology, which is already attracting strong customer interest.

The AI engine

The company sits at the centre of one of the largest investment cycles in the semiconductor industry’s history.

Many of the world’s leading AI processors, including Nvidia’s GPUs and much of the custom AI silicon designed by Amazon, Google and Microsoft, are manufactured by TSMC in Taiwan.

At the company’s April earnings presentation, CEO Che-Chia Wei described AI demand as “extremely robust”, driven by the shift from chatbots that answer questions to agentic AI systems capable of taking actions.

That transition requires significantly greater computing power, increasing demand for the advanced chips TSMC manufactures.

Advanced technologies, defined as chips produced using process technologies of seven nanometres or smaller, accounted for 74% of wafer revenue in the first quarter.

TSMC’s three-nanometre technology alone contributed 25% of wafer revenue.

Reports have indicated that Nvidia has reserved roughly 60% of TSMC’s advanced chip-packaging capacity for 2026, highlighting continued supply constraints across the AI semiconductor market.

South Korean chipmaker SK hynix, known for its high-bandwidth memory chips, is preparing to raise roughly $28 billion (€24.5bn) on Wall Street, a sum surpassed only by SpaceX’s record flotation last month.

ADVERTISEMENT

ADVERTISEMENT

It is an extraordinary outcome for a firm that once survived on job cuts and asset sales.

Pricing is due on Thursday, with trading expected to begin on Friday under the ticker SKHY.

SK hynix is issuing 17.79 million new shares in the form of American depositary receipts (ADRs), each representing a tenth of a Seoul-listed share, and cornerstone investors including Baillie Gifford and funds run by Coatue Management have signalled interest in up to $7 billion (€6.1bn) worth of stock.

The target was trimmed from an initial $29.6 billion (€25.9bn) after the shares slipped in recent weeks.

ADRs are certificates traded on a US exchange that stand in for shares held abroad, letting American investors buy into a foreign company without dealing in a foreign currency or market.

Unlike a conventional flotation, this is not SK hynix’s stock market debut. Its primary listing remains on Seoul’s Kospi index, and the Nasdaq offering simply opens a second, dollar-denominated avenue for investors to gain exposure.

The listing arrives with the company already worth more than $1 trillion (€876bn), a threshold also crossed by rivals Samsung Electronics and Micron, after a surge of more than 200% this year.

Proceeds will fund new fabrication plants, chiefly a vast cluster in Yongin, plus its first US packaging facility in Indiana.

The move is partly about valuation. Korean-listed chipmakers have long traded at a discount to American peers, and a Nasdaq listing offers a chance to close that gap.

The AI memory boom — and the risks

The AI build-out has transformed the industry’s economics.

As hyperscalers pour hundreds of billions into data centres, memory prices have exploded, with DRAM up 44% and NAND flash up 53% in a single quarter, according to Citi Research, and manufacturers have already sold most of their 2026 production.

SK hynix reported first-quarter revenue above 50 trillion won (€29bn) and operating margins north of 70%, figures unheard of for a chipmaker, and commands about 60% of the high-bandwidth memory (HBM) market, according to Counterpoint Research.

Yet the timing is delicate.

Memory has always been a brutally cyclical business. The AI-driven rally that transformed SK hynix has begun to wobble as chip stocks sold off sharply across Asia last week, and Samsung lost more than $100 billion (€87.5bn) in market value despite posting a record profit.

Investors are increasingly asking whether the vast sums being spent on AI infrastructure will earn a return, a question that the Bank for International Settlements raised in late June when it warned that the boom could seed the next financial crash.

Built, broken and rebuilt

Those concerns are not new for SK hynix.

SK hynix traces its roots to Gukdo Construction, founded in 1949, which moved into electronics in 1983 as Hyundai Electronics, an arm of the Hyundai empire.

The Asian financial crisis of the late 1990s brought disaster. Under an IMF-backed restructuring of the Korean economy, Hyundai absorbed rival LG’s semiconductor business, creating a giant that promptly buckled under its own debts.

Salvation came in stages.

Renamed Hynix Semiconductor in 2001, a contraction of “high” and “electronics”, the firm cut jobs, shed assets and split from Hyundai. Profits returned, but the violent swings of the DRAM market left it perpetually exposed.

Starved of capital, it was rescued in 2012 by the telecoms conglomerate SK Group, becoming SK hynix. The takeover proved decisive. SK Group poured money into high-bandwidth memory, then a costly and unprofitable technology that few believed in.

Today it has become the scarcest commodity in AI computing. And the firm employs nearly 46,900 people.

Additional sources • AFP

The second half of the year rests on a delicate chain of dominoes, according to a new briefing from Oxford Economics, and whether the US-Iran peace agreement holds is the factor that determines how the rest fall.

ADVERTISEMENT

ADVERTISEMENT

“Its durability will determine whether the global economy gets an energy-driven disinflation tailwind or absorbs a second oil shock,” stated chief global economist Ryan Sweet in the report, calling the deal “the key domino that will determine whether other risks are amplified or dampened”.

The consultancy expects the global economy to accelerate, forecasting annualised growth of 3.1% in the second half against an estimated 1.6% in the first, powered chiefly by cheaper oil feeding through to household incomes, although Sweet puts the odds of reaching a durable deal at “a coin flip”.

If the truce holds, Oxford Economics sees Brent crude averaging in the low $70s per barrel, easing inflation and financial conditions across emerging markets and tech valuations.

If it breaks, the consequences would not stay contained to the oil market.

Early on Wednesday, the US military attacked Iran after it said Tehran struck three ships in the Strait of Hormuz. Iran retaliated with strikes targeting Bahrain and Kuwait. The regional crossfire raised the risk that the interim agreement to halt fighting in the war could break down. However, the exchange of fire followed a pattern of similar attacks during the deal’s shaky ceasefire, and neither country immediately signalled it would step away from the negotiating table.

Oil prices reacted to the attacks by increasing more than 3% by Wednesday morning, with international benchmark Brent trading above $76 a barrel.

“A peace deal breakdown won’t just raise oil prices, it would also increase pressure on AI supply chains in Asia, force central banks to be hawkish, tighten financial conditions, and could shift the outcome of the US midterms and Israeli elections […] the cascade runs fast,” Sweet stated.

A coinflip with a $20 spread

Not everyone shares Oxford Economics’ outlook for oil prices.

Morgan Stanley’s mid-year outlook, published in May, forecast crude climbing back to roughly $90 a barrel by the end of the year, a gap of some $20 compared with Oxford Economics’ forecast that amounts to two different bets on the same peace process.

The World Bank is also more cautious, forecasting Brent crude to average about $94 a barrel this year while warning that global GDP growth will slow to 2.5% in 2026.

Reflecting on how the recent exchange of attacks is testing the fragile truce, Sweet said, “Traffic through the Strait of Hormuz is a good bellwether. The deal committed to fully restoring traffic through the chokepoint within 30 days, making mid-July the first hard deadline,” he explained.

“A sustained return to 75% or more of pre-war traffic by mid-July would increase the odds that the agreement is holding and vice versa,” Sweet concluded.

The other indicator, he says, is whether Iran formally invokes the accord’s Lebanon clause over Israeli strikes, and whether its response comes in military or rhetorical form.

Tariffs, trade and AI

Trade is another risk that could reshape the outlook.

US Section 122 tariffs are due to expire on 24 July, but Washington has already lined up replacement levies under Section 301. Oxford Economics expects the changes to push effective tariff rates higher from late July as the US seeks to maintain monthly tariff revenues of between $25 billion (€21.8bn) and $30 billion (€26.2bn).

Europe is also taking a tougher stance. The European Commission has more than 50 trade-defence investigations open against China, up from 17 a year ago, and plans to unveil a broader economic security strategy by September.

These trade tensions also feed into the AI boom that has powered financial markets this year.

Oxford Economics notes the US AI industry depends heavily on semiconductors and other hardware shipped from Northeast and Southeast Asia, the regions with the most to lose from any further disruption to commodities passing through the Strait of Hormuz.

Meanwhile, the Bank for International Settlements (BIS), the umbrella body for central banks, warned that the AI boom increasingly rests on opaque “circular financing” between chipmakers, cloud giants and artificial intelligence labs, as well as lightly regulated private credit, where lending to the sector has quadrupled in five years.

The BIS’s Asia-Pacific chief, Zhang Tao, cautioned that the sector’s reliance on non-bank funding means an AI downturn could trigger a sharper and faster correction than a traditional banking crisis.

Sweet modelled what such a reversal could look like.

“We have created a so-called tech bust scenario where US technology stocks fall by 25% over the course of a year,” he told Euronews.

According to Sweet, such a shock would cause the US economy to “grind to a halt”, spilling over to technology exporters and investor sentiment worldwide, leaving global growth 1.1 percentage points below Oxford Economics’ baseline next year.

Central banks, ballots and the calendar

The final dominoes are policy and politics.

Oxford Economics expects the major central banks to prove more dovish than financial markets currently anticipate, though they could pivot quickly if traffic through the Strait of Hormuz falters or AI-input prices signal supply stress.

The nearest test is the Federal Reserve’s rate decision under chair Kevin Warsh later this month, coming on the heels of June’s soft jobs report.

Beyond that lie November’s US midterms and Israel’s general election, due by late October, both of which could influence the Middle East peace process. In September, German state elections could also test the coalition behind Germany’s fiscal policy, a key driver of the eurozone economy.

Oxford Economics also flags genuine upside, from stronger AI-driven productivity to an EU economy that weathered the second quarter surprisingly well.

Whether the resilience in Europe is real will show up first in Germany and in credit data, Sweet argues.

“If corporates were absorbing margin compression from the jump in energy prices without cutting investment and drawing down credit lines, that would strengthen the case that underlying momentum in the economy is better than we expected,” he told Euronews, adding that a contraction in eurozone bank lending would push the other way.

It is important to highlight that the typical Oxford Economics forecast miss is nearly a full percentage point, and the range around this assessment in particular is wider than usual.

Published on 07/07/2026 – 11:28 GMT+2•Updated
11:34

The South Korean technology giant Samsung said on Tuesday it expects operating profit of about 89.4 trillion won (€51bn) for the April-June quarter, roughly nineteen times the 4.7tr won (€2.7bn) it earned a year earlier and more than it made in the previous three years combined.

ADVERTISEMENT

ADVERTISEMENT

The extraordinary numbers reflect the same force reshaping the memory industry worldwide: the race to build AI data centres has pushed chip prices to record highs.

According to Citi Research, average selling prices for DRAM memory rose 44% quarter on quarter, and NAND flash 53%, as AI demand spilled beyond specialised high-bandwidth memory into the conventional chips that go into phones, servers and PCs, with customers now chasing longer-term supply contracts.

The estimate beat analyst forecasts, but far from celebrating, the market sold.

Samsung shares fell by over 10% before closing nearly 7% lower, dragging rival SK Hynix and the wider Kospi index down with them.

Samsung’s stock has more than doubled this year alone, so a historic quarter was already priced in, and leveraged local ETF products tracking the shares have made them prone to outsized moves.

There was also a blemish in the numbers as revenue of 171tr won (€97.6bn), though up 129% year on year, came in slightly below forecasts.

“We believe the slight revenue miss was largely driven by more moderate DRAM price hikes than expected, which likely spooked investors who are increasingly pricing in structural strength in memory prices,” said Jing Jie Yu, an analyst at Morningstar.

Hanging over everything is durability.

Investors are increasingly asking whether the technology giants bankrolling the AI build-out can sustain their spending without piling up debt against a payoff that remains unproven, the worry behind last week’s chip sell-off across Asia.

Samsung publishes its full results, with a breakdown by division, on 30 July, a report the market will scour for clues about whether the boom is structural or simply another memory cycle nearing its peak.

Additional sources • AP

Published on
06/07/2026 – 11:38 GMT+2

The South Korean government intends to set aside the extra tax income flowing from its record-breaking chip industry in a dedicated “future response fund”, the presidential office said, using the proceeds of the AI boom to bankroll public projects ranging from industrial infrastructure to support for younger generations.

ADVERTISEMENT

ADVERTISEMENT

Behind the windfall sit Samsung Electronics and SK hynix, whose memory chips have become essential to the data centres powering the global AI race.

Their record profits this year have propelled the wider economy, and swollen the government’s tax receipts along the way.

Presidential chief of staff Kang Hoon-sik outlined the plan at a meeting between the government and the ruling party on Sunday, saying the fund would help finance large-scale projects built around AI and semiconductors, while also tackling inequality and helping young people with housing, start-ups and work.

Kang warned that the extra revenue thrown off by the chip boom must not be squandered at what he described as a decisive moment for the country’s future.

No figure was provided for the fund’s size, as the government will consider its use at a fiscal strategy meeting this month before consulting the public.

In an interview with the Dong-A Ilbo newspaper, Kang added that part of the money would go towards the utilities on which chip plants depend, above all power and water.

A boom that keeps giving

The windfall reflects an extraordinary run for Korea’s chipmakers.

Samsung shares surged more than 170% in the first half of the year, and SK hynix shares rose more than 300%, carrying both companies past $1 trillion (€874bn) in market value.

Samsung is due to publish preliminary second-quarter earnings on Tuesday, while SK hynix plans to raise 45 trillion won (€25.7bn) through a listing on the Nasdaq.

Both are also part of an 800 trillion won (€457bn) public-private push, unveiled last week, to build a new chipmaking hub in the country’s southwest.

How the windfall should be spent has become a live political debate.

In May, presidential policy chief Kim Yong-beom floated using it for start-ups, young people, basic income schemes in rural and fishing communities, and support for artists.

The boom has also emboldened workers as Samsung averted a major walkout in May by agreeing to a bonus deal with its largest union.

Additional sources • AFP

Published on 03/07/2026 – 10:07 GMT+2•Updated
10:11

Stock markets across Asia mostly advanced on Friday, taking their cue from a fresh record close for the Dow on Wall Street, as some of the AI-linked shares battered in this week’s sell-off found their feet again while others kept falling.

ADVERTISEMENT

ADVERTISEMENT

The volatility was calmer than the heavy selling seen a day earlier, when worries about stretched technology valuations sent semiconductor shares tumbling across the region.

At the time of writing, South Korea’s Kospi led the bounce, climbing over 4% to recoup part of the nearly 8% plunge it suffered on Thursday. Samsung Electronics, the country’s largest company and a major chipmaker, jumped 7%, while smaller memory rival SK Hynix rose 4.9%.

In Tokyo, the Nikkei 225 added 1%, helped by a 6.6% leap in memory maker Kioxia, although chip-equipment supplier Tokyo Electron slipped 2.5%.

Elsewhere, Hong Kong’s Hang Seng gained 1.7% and the Shanghai Composite rose 0.7%, while Australia’s S&P/ASX 200 advanced 1.3% and Taiwan’s Taiex bucked the trend, easing 0.6%.

As for European markets, both the Euro Stoxx 50 and the broader pan-European Stoxx 600 opened within a 0.3% range.

The UK’s FTSE 100, Germany’s DAX 30, France’s CAC 40 and Italy’s FTSE MI, all traded between 0.1% and 0.3% higher.

Spain’s IBEX 35taly’s FTSE MIB led the pack and rose about 0.4%.

Wall Street’s record, a cooler jobs report and oil

US stocks were mixed on Thursday, but the Dow still managed a fresh peak, rising 1.1% to 52,900.

The broader S&P 500 ended virtually flat despite seven in ten of its members gaining, held back by another retreat in chip stocks, while the technology-heavy Nasdaq fell 0.8%.

Sentiment drew support from data showing US employers added 57,000 jobs last month, well below the 100,000 forecast and a slowdown on May.

A softer labour market could ease inflation pressure and, with oil back below its pre-war levels, may lessen the case for the Federal Reserve to raise interest rates repeatedly this year, an outcome investors would welcome.

Crypto-linked shares also firmed as Bitcoin rose about 2%, lifting Robinhood and Coinbase alongside it.

Still, the AI trade remained under strain.

Micron gave up an early gain to fall 5.5%, a day after a 10.6% slump, while Lam Research sank more than 10% and Nvidia, now worth close to $4.7 trillion, edged 1.4% lower.

On oil, Brent crude, the international benchmark, rose 1% to around $73 a barrel early Friday, while US crude added 0.5% to about $69, with prices still sitting below where they were before the Iran war began in late February.

Additional sources • AP

Published on
02/07/2026 – 9:39 GMT+2

Most Asian stock markets dropped on Thursday, dragged down by a wave of selling in semiconductor shares, as European bourses made a subdued start and Wall Street looked set to open in the red before the release of key US employment figures.

ADVERTISEMENT

ADVERTISEMENT

The pullback centred on the technology sector, where investors retreated from the chip stocks that have powered much of this year’s rally, amid growing unease that the vast sums Big Tech is spending on AI could leave the market awash with supply.

South Korea’s Kospi bore the worst of it, tumbling around 5% as its heavyweight chipmakers slid. Memory specialist SK Hynix lost close to 8% and Samsung Electronics fell more than 6%.

In Tokyo, the Nikkei 225 shed about 1.5%, with chip-equipment maker Tokyo Electron down around 5.6%, while Taiwan’s Taiex slipped 1.1% as TSMC, the world’s largest contract chipmaker, gave up 1.8%.

The falls followed a rough session for chip stocks on Wall Street this Wednesday, where Micron Technology dropped more than 10% and Intel sank around 9%.

The moves stand in sharp contrast to a stellar year for Asian tech, with the Kospi and the Nikkei still up roughly 85% and 34% respectively in 2026.

On the other hand, Hong Kong’s Hang Seng rose about 0.8%, lifted by an 8.7% jump in electric-vehicle maker BYD after it reported a second straight monthly rise in sales, while India’s Sensex added 0.5%.

In Europe, markets opened flat as both the Euro Stoxx 50 and the broader pan-European Stoxx 600 traded within a 1% range at the start of Thursday’s session.

The UK’s FTSE 100, Germany’s DAX 30, France’s CAC 40 and Spain’s IBEX 35, all traded between 0.1% and 0.3% higher.

Italy’s FTSE MIB led the pack and rose about 0.4%.

Oil extends its slide and US jobs in focus

Crude prices fell again, trading below where they sat before the Iran war began in late February, as hopes grew that supplies through the Strait of Hormuz will steadily recover.

Brent crude, the international standard, eased around 1% to about $70.89 a barrel while WTI, the US benchmark, dropped 3% to roughly $69.

Attention now turns to the US, where stock futures edged lower ahead of the June employment report, brought forward a day because of Friday’s Independence Day.

Economists polled by Dow Jones expect around 115,000 jobs were added last month.

The figure carries extra weight under the new Federal Reserve chair, Kevin Warsh, with investors wary that a strong reading could harden the case for keeping interest rates higher for longer.

According to economists at Capital Economics, demand for AI may keep growing but at a slower pace than many expect, a caution that helped sour sentiment towards the sector.

Additional sources • AP

For more than three years, the ‘Magnificent Seven’ or ‘Mag 7’, which includes Nvidia, Apple, Microsoft, Alphabet, Amazon, Meta and Tesla, carried Wall Street.

ADVERTISEMENT

ADVERTISEMENT

Then came June 2026.

Nvidia dropped over 5%, Microsoft fell about 17%, its worst monthly performance since December 2000, Alphabet declined nearly 6%, Amazon lost roughly 12% and Meta dropped around 11%.

As for Apple and Tesla, the companies had directionally different but equally volatile monthly moves.

Apple made a new all-time high closing price of $315.2 on the second day of the month but subsequently declined more than 10% from that peak.

On the other hand, Elon Musk’s company dropped more than 6% in the first week of June but clawed most of that back by the close of the month, ending roughly flat.

Taken together, the ‘Magnificent Seven’ erased about $2.3 trillion (€2tn) in market value in a single month.

What made the selloff remarkable was its breadth. Usually one or two stocks stumble while the others hold up. This time, nearly every member of the group moved lower.

The Roundhill Magnificent Seven ETF (MAGS), which holds all seven companies, fell about 13% from its late May record high.

So what happened to Wall Street’s favorite technology stocks? And why are investors backing away?

Growing pains and spending

The MAGS ETF bled more than $700 million (€615mn) over the month, its worst outflow since it launched in 2023, according to TradingView data. For a fund that had become the simplest way to bet on the US tech boom, the reversal was striking.

One name outside the club had it even worse. Oracle, a hyperscaler not included in the ‘Magnificent Seven’, crashed around 35%, its steepest month since September 1990, after alarming investors with a surge in AI spending and debt.

The fall wiped roughly $100 billion (€87.9bn) off the fortune of co-founder and billionaire Larry Ellison. The market punished the biggest AI spenders, and the numbers explain it.

The five largest hyperscalers are set to spend more than $700 billion (€615bn) on AI infrastructure this year. Microsoft alone is heading towards roughly $190 billion (€167bn), according to estimates from the Bank of America.

The bank said that hyperscaler capital spending has jumped from about 70% of operating cash flow in 2025 to nearly 100% in 2026.

The translation is simple: far less capital left over for share buybacks and dividends, and an increasingly larger bill that will need to be justified with future revenue as costs are climbing too.

The ‘Magnificent Seven’ are the biggest buyers of the memory that feeds AI data centres, and those chips have become scarce and expensive.

Micron Technology, one of the main memory chipmakers, reported earnings per share of $24.67 for its latest quarter, up from $1.68 a year earlier, close to a fifteenfold jump.

Prices for DRAM, the memory inside almost every device, rose as much as 98% in the first quarter alone, a surge some in the industry have nicknamed “RAMageddon”.

A quieter shift beneath the surface

While the biggest technology stocks struggled, the rest of the market continued to rise.

LPL Financial chief equity strategist Jeff Buchbinder points to that trend. Excluding the ‘Magnificent Seven’, the remaining S&P 500 companies grew earnings by 17.5% in the first quarter, helped in part by semiconductor and memory producers.

Buchbinder expects that figure to exceed 20.5% in the second quarter. Meanwhile, the earnings growth projection for the ‘Magnificent Seven’ will be lower than that.

In other words, the other 493 companies are now growing earnings faster than the market’s biggest stars, and investors have noticed.

By late June, the S&P 493 – which excludes the ‘Magnificent Seven’ – had climbed 13.7% for the year. In contrast, the ‘Magnificent Seven’ basket was down 6.6%, while the broader S&P 500 posted a more modest 7.4% gain.

According to veteran investor Ed Yardeni, investors are beginning to show signs of AI fatigue, questioning whether unprecedented spending on infrastructure will ultimately generate attractive returns as cheaper open source models proliferate and AI token prices continue to decline.

Are the ‘Magnificent Seven’ still “magnificent”?

The ‘Magnificent Seven’ still delivered an estimated 29% earnings growth in the first quarter, and they are unlikely to lose their leadership positions anytime soon.

Yet, the debate has shifted.

Investors are no longer asking whether AI will transform the economy. They are asking when hundreds of billions of dollars in AI investment will begin producing meaningful returns.

June may have offered the first clear answer.

The AI trade is no longer a one way bet on seven companies. The ‘Magnificent Seven’ created the AI boom, but they are no longer the only way to invest in it.

In its Annual Economic Report, published on Sunday, the Bank for International Settlements (BIS), known as the central bank for central banks, warned that the enormous spending on AI is accumulating financial vulnerabilities that could amplify any future shock and spread from markets into the wider economy.

ADVERTISEMENT

ADVERTISEMENT

Presenting the findings, BIS general manager Pablo Hernández de Cos said the message was one of “urgency”, with policymakers urged to act before any reversal makes the eventual adjustment more painful.

At the core of the warning is the scale of the spending, despite massive investment having supported global growth over the past year.

The five largest “hyperscalers”, the technology giants racing to build AI infrastructure, are on track to commit more than $1 trillion (€878bn) to AI-related investment across 2025 and 2026, a pace that is outstripping their earnings and free cash flow and pushing some to borrow heavily to keep up.

The BIS suggests this race is fuelled by a belief that only a handful of dominant players will ultimately prevail, encouraging firms to pour money into projects whose returns remain deeply uncertain.

Echoes of past manias

The report sets today’s AI boom against a long historical lineage, from the canal mania of the 1830s and Britain’s railway mania of the 1840s to the electrification of the 1920s and the dotcom bubble.

Each began with a genuine technological breakthrough that attracted more capital than commercial returns could justify, the BIS notes, with each episode ending “with an eventual reversal in investment, inducing economy-wide recessions”.

Compounding the danger are stretched share prices and opaque financing.

The BIS highlights the spread of “circular financing”, in which chipmakers and cloud giants take equity stakes in AI labs that then commit to buying their chips and computing power, effectively recycling money back to the original investors as revenue.

Much of the funding now flows through hedge funds and private credit vehicles that face lighter scrutiny than banks.

According to Zhang Tao, the BIS chief representative for Asia and the Pacific, that reliance on non-bank channels means an AI downturn could unwind into a sharper, faster crash than a traditional banking crisis.

The hidden costs of data centres

Beyond financial markets, critics argue the true cost of the AI build-out is being obscured in plain sight.

A central concern, examined by the Wall Street Journal, is how the technology giants account for their data centres.

By assuming the expensive equipment inside them will stay useful for longer, firms can spread its cost over more years, lowering the depreciation charged against profits in any given period and making earnings look healthier than the underlying cash burn implies.

However, the specialist chips at the heart of these facilities may become obsolete far faster than those extended schedules assume, leaving a gap between reported profits and economic reality, as well as a balance sheet more exposed than it appears should demand disappoint or a sizable need to replace hardware arise.

The physical scale is staggering.

Columbia University economist Stijn Van Nieuwerburgh estimates the build-out could cost in the region of $8 trillion (€7tn) over the next six years, financed in part through the kind of off-balance-sheet arrangements the BIS flagged.

The costs are also no longer confined to corporate accounts.

Some economists now warn of a so-called “third wave” of inflation, after the pandemic and tariffs, driven this time by the AI build-out. As chip manufacturers prioritise high-margin parts for AI servers, the resulting squeeze on memory and storage has rippled out to consumer electronics.

For example, Apple raised prices on its MacBooks, iPads and other devices last week, citing an “extraordinary surge in demand for memory and storage” and saying it had “never seen a component price increase this much, this quickly”.

The company’s shares fell around 6%, their worst day in over a year, as Microsoft, Nintendo and Sony have also made similar moves.

Beyond hidden costs and inflationary pressures, where the strain may spread furthest is raw power.

Goldman Sachs expects data centres to account for nearly half of the growth in US electricity demand by 2030, with consumer power prices forecast to rise around 6% a year through 2026 and 2027.

The BIS itself notes that the build-out’s hunger for electricity is already pressuring prices and input costs, with potential spillovers to inflation, though it stresses, as do many economists, that AI could yet prove disinflationary if its promised productivity gains eventually arrive.

Halfway through a turbulent year, a clear pattern has emerged across global markets: anything tied to the physical build-out of AI has soared, while several other assets that investors traditionally turn to in uncertain times have stumbled.

ADVERTISEMENT

ADVERTISEMENT

War in the Middle East, political upheaval and an oil-price spike formed the backdrop, yet stock markets in several regions still pushed to fresh record highs.

According to Dan Coatsworth, head of markets at AJ Bell, companies on the receiving end of the AI spending boom were the standout investments of the first half, while Bitcoin proved “a shocker” and gold lost its shine.

It is, Coatsworth noted, a remarkable run of events for only half a year’s worth of trading.

The most spectacular gains came from an unglamorous corner of the technology world: the firms that make memory chips.

As demand for AI computing collided with tight supply, prices surged and took shares with them. SanDisk led the US market with a gain of over 850% in six months, while Western Digital, Micron Technology and Seagate Technology all more than tripled in value, a pace of return that would ordinarily take many years to achieve.

The driver is the vast quantity of high-speed memory and storage needed to train and run AI systems as the largest technology companies race to expand their data centres.

Other US equities that soared on the back of the AI trade include Intel, Dell, Advanced Micro Devices (AMD) and Applied Materials, which all rose between 150% and 280% year to date.

The rush also lifted emerging markets, where Asian chipmakers such as TSMC and SK Hynix carry heavy weight, helping South Korea’s KOSPI double in value, Japan’s Nikkei 225 climb roughly 40% and the MSCI Emerging Markets index rise by around 27%.

In Europe, the FTSE 100 gained 7% in the first half of the year, France’s CAC 40 rose 5%, while Germany’s DAX gained 2%. Meanwhile, the MSCI India index fell 5% and Hong Kong’s Hang Seng lost 6%.

Notably, the memory rally has begun to unwind in recent days, with several of the same names caught in a sharp technology sell-off.

The fallen favourites, takeovers and the trades that cooled

The flipside was brutal for yesterday’s winners.

Previous AI darlings Meta and Microsoft were left behind, down 14% and 24% respectively on a total-return basis, as heavy AI spending turned the technology giants into more capital-hungry businesses and investors stopped paying a premium for them.

Microsoft now trades at its cheapest level in a decade, leaving both it and Meta valued more modestly than McDonald’s, an outcome few would have predicted at the height of the “Magnificent 7” craze.

Elsewhere, the assets many expected to lead disappointed.

Gold took investors on a volatile ride. After surging to a record high of $5,594.82 an ounce on 29 January, the precious metal lost around 28% from its peak despite the geopolitical turmoil that would normally send investors flocking to safe-haven assets. Instead, its appeal was undermined by higher bond yields and cash rates, which offer an income that a gold bar cannot.

Bitcoin fared worse still, falling 28% since the start of the year as enthusiasm for crypto drained away and money rotated towards technology shares instead.

In the UK, takeovers did much of the heavy lifting.

Six FTSE 100 companies, among them Glencore, Schroders and Segro, attracted bid interest in the first half, a sign that buyers still see value in British blue chips even after a three-year re-rating.

Housebuilders such as Persimmon struggled against a sluggish property market, while tech-adjacent names like Experian and RELX were swept up in fears about AI disruption.

One trade that conspicuously cooled was defence.

After a storming 2025, the likes of BAE Systems, Germany’s Rheinmetall and America’s Palantir all gave ground, as the good news on rising military budgets looked fully priced in and investors drifted elsewhere.

This article does not constitute financial advice. Always do your own research and invest according to your specific circumstances.

Published on
25/06/2026 – 10:43 GMT+2

Micron, one of only a handful of companies able to make advanced memory chips at scale, said on Wednesday that revenue in the third quarter reached $41.4 billion (€36.5bn), more than four times the $9.3 billion (€8.2bn) it recorded in the same period last year.

ADVERTISEMENT

ADVERTISEMENT

The figure also comfortably beat the roughly $35.7 billion (€31.4bn) analysts had forecast, while profit climbed even more dramatically.

The Idaho-based group posted net income of $28.24 billion (€24.9bn), or $24.67 per share, against less than $2 billion (€1.7bn) a year ago. Adjusted earnings of $25.11 a share sailed past the $20.49 expected.

The market reaction to the impressive results was immediate.

Micron shares rose more than 15% in after-hours trading to around $1,213, leaving the company valued at roughly $1.16 trillion (€1tn).

The stock has now climbed about 700% over the past year, one of the most dramatic re-ratings of any large company through the AI boom, reflecting a fundamental shift in the economics of the AI build-out.

The vast data centres being constructed by hyperscalers such as Amazon, Microsoft, Google and Meta, which have collectively earmarked hundreds of billions of dollars in capital spending this year, depend on enormous quantities of high-bandwidth memory, a specialised chip that sits alongside the processors made by Nvidia and others.

Micron has said its entire 2026 output of these chips is already sold out under fixed-price contracts.

According to CEO Sanjay Mehrotra, the results reflect what he called the strategic value of memory in the AI era.

The company pointed to a series of multi-year customer agreements that it expects to make earnings more durable and predictable, a notable claim in an industry long defined by brutal boom-and-bust cycles.

Margins to rival the biggest names

What has startled analysts most is Micron’s profitability.

The company reported a gross margin of around 85% for the quarter, a level that now rivals or exceeds those of far larger technology names such as Nvidia and Meta, an extraordinary position for a memory maker historically squeezed by volatile chip prices.

The tightness of supply, with new factories not expected to add meaningful output until 2028, has handed producers exceptional pricing power.

Micron’s guidance was more striking still.

The company expects revenue of around $50 billion (€44bn) in the current quarter and adjusted earnings of roughly $31 a share, implying the boom is accelerating rather than fading. It is ramping up investment to match, lifting planned capital spending to about $27 billion (€23.7bn) this fiscal year and signalling a further jump in 2027, management told analysts during the earnings call.

The results offer reassurance to investors betting that AI infrastructure spending remains robust, with Micron’s order book serving as a real-time gauge of that demand.

The open question, as ever in the memory industry, is how long the upswing can last before supply catches up. Even the most bullish observers acknowledge that risk has not completely disappeared.

HARRISBURG, Pa. — Six months after President Trump warned states not to regulate artificial intelligence, they are increasingly doing just that.

Congress has stalled on producing federal regulations of artificial intelligence as states forge ahead and scrutinize how chatbots interact with children, how AI systems are used by employers and what developers must do to try to prevent an AI-caused catastrophe.

State lawmakers have stepped back from earlier, wider-ranging attempts to regulate AI that were vetoed or otherwise derailed by governors who viewed the measures as too onerous toward the industry’s development, including efforts to hold developers accountable for bias in AI systems.

But they are returning with legislation that is more targeted and, often, probes the corners of life where Americans interact with AI but may not know it.

Presidential power versus state power

Trump’s move to restrain states’ actions on AI drew criticism from members of both political parties and civil liberties and consumer rights groups who worried that banning state regulation would amount to a gift to AI giants, who enjoy little to no oversight.

Trump has made AI a top national and economic security priority, and he said that letting states clutter the regulatory playing field for an industry that’s spending trillions of dollars and driving the economy is too risky in the race with China for AI superiority.

Trump issued an executive order that directed the attorney general to create a task force to challenge state laws that are more than “minimally burdensome,” and directed the Commerce Department to draw up a list of problematic regulations. It also threatened to restrict funding from a broadband deployment program and other grant programs to states with AI laws.

The White House said it wouldn’t target state laws that seek to prevent fraud and protect consumers and children.

In the meantime, the Trump administration released a “national policy framework” in which it urged Congress to preempt state AI laws that are out of step with its regulatory worldview and to pass legislation to protect children, intellectual property rights and free speech. A recent bipartisan draft proposal in the House was met with withering criticism from key Democrats and Republicans.

The White House has given no indication that it has made good on its threat to enforce the president’s executive order by going to court against a state’s AI law or withholding money. In a statement, it said the Trump administration is “eager to work with partners” to enact its policy framework.

States seem largely unrestrained by Trump

Trump’s executive order didn’t seem to discourage states from trying to regulate how AI is used. More bills have been introduced this year than last, including by Republicans, said Justine Gluck, policy director of the Future of Privacy Forum, a nonprofit that advocates for data privacy in technology and whose members are from industry, academia and civic groups.

In Illinois, legislation on the desk of Democratic Gov. JB Pritzker piggybacked on elements of laws passed last year in California and New York that require developers of large advanced AI models to create protocols to prevent their systems from causing catastrophes such as a biological weapons attack, power outage or large-scale hack.

Illinois added a requirement that AI developers must get an independent auditor to review whether they are complying with their own policies. Analysts see it as a step toward requiring AI developers to take greater accountability for their products.

The bill’s sponsor, Democratic state Sen. Mary Edly-Allen, brushed aside Trump’s threat.

“I don’t know if you’ve met Illinois, but we’re pretty independent,” Edly-Allen told the Associated Press.

The bill drew nearly unanimous support, signaling a willingness by members of Trump’s party to cooperate with Democrats in filling the AI regulatory vacuum left by the federal government.

This kind of legislation is expected to expand to other states.

Regulating chatbots, especially for children

A growing number of states are imposing restrictions on how AI chatbots can interact with people, especially children. A mix of Republican- and Democratic-led states have passed such laws this year, including Colorado, Connecticut, Idaho, Iowa, Nebraska and Oregon.

In many cases, states want companies to tell people when they are interacting with AI instead of a human. Many want chatbots to be restricted in how they interact with minors, parents to have control over their child’s access, and data given to chatbots to be kept private.

In recent weeks, Connecticut enacted provisions for companion chatbots that sustain an ongoing relationship with a human. Under them, a chatbot must not be able to interact with someone under 18 unless it is programmed against encouraging self-destructive behavior and provides parents with tools to manage the child’s use.

Transparency in AI and decision-making

In California, lawmakers are advancing the “No Robo Bosses Act of 2026” to prohibit employers from relying solely on AI to fire or discipline workers, and an expansion of how the state regulates AI chatbots, including banning chatbot outputs to children from being used for advertising.

Colorado in May required companies that deploy AI systems in important areas such as employment, education, housing or banking to tell people when AI is being used to influence a decision made about them.

It was a stab at regulating what researchers say is the bias inherent in AI systems that sort through a consumer’s data and render consequential decisions — including who gets hired, a home loan or medical care. But it watered down a 2024 law aimed at preventing AI’s penchant to discriminate, amid pressure from Democratic Gov. Jared Polis.

In Connecticut, lawmakers required employers who are using employment-related AI systems to tell employees or job applicants that they are interacting with AI.

Meanwhile, Connecticut, Washington and Utah required AI developers to embed data into digital content that will allow users to determine whether the content — such as photos or video — has been created or altered by AI.

More laws are possible this year.

Some Republican-led states hold back

In Florida, the state House refused to advance what Republican Gov. Ron DeSantis called his AI “Bill of Rights” legislation. It included provisions to give parents control over their children’s access to companion chatbots and to require companies that use chatbots to tell consumers when they are interacting with AI instead of a human.

Florida House Speaker Daniel Perez, a Republican, said Trump had made it clear that the federal government should be in charge of AI regulation. DeSantis panned that idea, noting that the federal government isn’t acting.

In Utah, progress stalled on legislation modeled on laws in New York and California after the White House sent a one-sentence memo to lawmakers there to warn that it was “categorically opposed” to the bill.

Levy writes for the Associated Press.

WASHINGTON — In congressional races across the country, a new crop of super PACs is taking to the air with millions of dollars worth of advertisements to sway voters.

“President Trump said it best, ‘Celeste Maloy will never let you down,’” says one advertisement supporting the Utah Republican representative in her upcoming primary election.

“Standing up to big pharma, fighting for local jobs, Val Hoyle doesn’t back down,” says an ad backing the Oregon Democratic representative ahead of her primary victory last month.

The super PACs have nondescript names — such as Jobs and Democracy PAC and American Mission — and the text is so generic that it almost seems to have been created by artificial intelligence.

That isn’t so far off the mark. The AI industry has funded the ads.

One network of super PACs is linked to Anthropic, maker of the popular AI tool Claude, and the other to Open AI, maker of ChatGPT.

They have been among the most prolific political spenders so far in the 2026 midterm elections, splashing out more than $37 million to date to influence races across the country and making the groups among the biggest outside spenders so far in congressional races. That number could grow exponentially as campaign season heats up closer to the November election — and as the Silicon Valley giants prepare initial pubic offerings that are poised to raise billions of dollars for the companies and their executives.

The AI political spending boom comes as emerging technology companies have become increasingly “comfortable with using their power to achieve a political goal,” said Adam Kovacevich, a former Google public policy executive and founder of Chamber of Progress, a technology trade group with a progressive orientation.

The leading AI companies have a history.

Anthropic was formed by former OpenAI employees who were concerned that the company was less focused on its original mission to safely harness the power of AI.

The companies are now the leading drivers of the burgeoning AI industry, and their competing views about how the technology should be regulated are playing out in a wide-ranging political ad spending war that has targeted congressional races in big cities and rural areas alike.

OpenAI thinks AI should be regulated solely at the federal level.

Anthropic calls for more stringent regulation and supports efforts by states such as New York and California that have passed more aggressive AI laws.

The groups spending in these races are super PACs, which are able to raise and spend unlimited amounts of money in federal races thanks to the 2010 Citizens United Supreme Court decision.

In some races, the AI-backed political groups have spent more than the candidates they are backing.

“There was no way as a grassroots person that I could compete with that kind of money,” said Al Olszewski, whose opponent in a Montana Republican congressional primary beat him by 30 points after getting a boost from $877,000 in ads from a super PAC backed by OpenAI’s co-founder. “I got crushed.”

The AI behemoths have emphasized that they are independent from the political groups.

One group counts $25 million in support from OpenAI co-founder Greg Brockman and his wife, Anna, alongside $100 million tied to one of Silicon Valley’s biggest venture capital firms, which holds a large stake in OpenAI. The global policy chief for OpenAI was reportedly involved in conceiving the group.

The other side has gotten $20 million from Anthropic and millions more from donors whose identities are not public.

This anonymous political cash is commonly known as dark money, and its prevalence is growing.

(Los Angeles Times photo illustration; source photos courtesy of the Tech Oversight Project)

“This has become very normalized now,” said Brendan Glavin, director of insights at OpenSecrets, which tracks campaign spending. “In 2024, we tracked over $1 billion in dark money.”

That total was $350 million higher than the previous presidential election.

The crypto playbook

The political activity of these AI companies and executives reflects a dramatic shift from how emerging technology companies have historically engaged with politics.

Google, for example, didn’t hire its first in-house Washington lobbyist until after the company had gone public in 2005.

“I think that for a long time, the tech industry lobbying strategy was just ‘leave us alone,’” Kovacevich said.

He sees the spending by these AI-linked super PACs as following the recent playbook developed by the cryptocurrency industry, which has funded the only network of political groups that has spent more on congressional races this year than those linked to OpenAI.

“I think what the crypto industry realized was that there’s no substitute for building up political power,” Kovacevich said.

The political stakes for these technology companies are significant.

“AI policy is far from settled,” said Asad Ramzanali, the former deputy director for strategy in the White House Office of Science and Technology Policy during the Biden administration and the director of artificial intelligence and technology policy at the Vanderbilt Policy Accelerator.

Earlier this month, the Trump administration banned foreign nationals from using the most powerful AI model developed by Anthropic — and even banned the company’s own employees from it — which forced the company to restrict access for all users.

Manhattan matchup

The two super PAC networks have largely shied away from producing ads that mention AI and have mostly chosen to avoid competing against each other in the same races.

There’s one big exception.

In the marquee Manhattan Democratic congressional primary to replace retiring Rep. Jerry Nadler (D-N.Y.), each side has spent millions of dollars.

While the field includes Kennedy scion and social media star Jake Schlossberg and former Republican turned Trump critic George Conway, the target of all the AI-backed spending has been Alex Bores, a former Palantir data scientist who now serves in the New York state Assembly.

New York congressional candidate sponsored a state measure Bores requiring major AI companies to be transparent about their safety protocols and promptly report safety incidents.

(Yuki Iwamura / Associated Press)

That’s because Bores sponsored a state bill, known as the RAISE Act, that requires major AI companies to be transparent about their safety protocols and promptly report safety incidents. The bill was signed into law in December 2025.

The ads sponsored by the group tied to OpenAI, which has spent more than $7.5 million in the race, paint Bores as someone who can’t be trusted.

They cite his support from other tech billionaires, including former crypto mogul and convicted financial fraudster Sam Bankman-Fried, whose super PAC spent $100,000 to support Bores in 2022 when he first ran for New York Assembly.

“Is that really who should be shaping AI safety for our kids?” one ad asks.

An ad sponsored by the Anthropic-backed network, which has also spent more than $7.5 million supporting Bores, makes the case that the bill he sponsored is exactly why he should be elected.

“As a computer engineer, Alex Bores saw how dangerous unregulated AI could be and he wrote New York’s RAISE Act to put real safeguards on A.I. and hold big tech accountable,” the ad says.

The AI ad barrage in New York has even included what might be considered a kumbaya moment in the ad wars — another super PAC created to support Bores is most heavily backed by both an employee of Anthropic and an employee of OpenAI, who both focus on AI safety.

The group, Dream NYC, has spent more than $1.7 million supporting Bores.

Bores and fellow New York State Assemblymember Micah Lasher have been atop the most recent polls in the race ahead of the June 23 primary.

A general view of businesses in St. George, Utah, on Wednesday.

(Ian Maule / For The Times)

Rural Republicans

For voters in many parts of the country, the debate over AI policy has played out locally as a debate over the massive data centers required to power the technology.

In Utah, a proposed data center in Box Elder County, backed by “Shark Tank” television personality Kevin O’Leary, has generated controversy because of questions about its impact on resources in the drought-prone state and its environmental effect on the nearby Great Salt Lake.

In the state’s most competitive Republican congressional primary — the vast, newly drawn 3rd Congressional District — both candidates expressed concerns about how the project has been developed and called for greater transparency in this plan and for future data centers in the state.

Utah congressional candidates Phil Lyman and Celeste Maloy in a debate on June 1. A super PAC backed by Anthropic has spent more than $920,000 to support Maloy.

(Rick Egan / Pool / The Salt Lake Tribune Via Associated Press)

Despite their similar position on the project, a super PAC backed by Anthropic has spent more than $950,000 to support Maloy, who is running in the new district after the boundaries of her old district changed.

“It’s a lot of money to throw at a race,” said her opponent, Phil Lyman, a former conservative Republican state Representative who ran to the right of Utah Republican Gov. Spencer Cox in an unsuccessful primary challenge in 2024.

Lyman insists he is no AI skeptic.

“I’m not anti data centers, I’m pro-transparency,” he said. “I think the future is bright with AI.”

The group said it is backing Maloy because it sees her as “someone who’s worked the issue” of AI regulation and who “has demonstrated leadership” with Republicans in Congress.

Maloy’s campaign didn’t respond to request for comment.

Utah congressional candidate Phil Lyman speaks during a Cottage Meeting at the SunRiver Community Center Ballroom in St. George, Utah, on Wednesday.

(Ian Maule / For The Times)

But Lyman suspects the group’s support for Maloy ahead of their June 23 primary has more to do with old-fashioned politics than any emerging technology.

One of the two co-founders of the political group is Chris Stewart, Maloy’s predecessor in Congress.

“Everything that they’re doing feels very coordinated,” Lyman said. “It makes you wonder if he’s still really controlling that seat.”

When Brian Grazer has an idea for a movie, he now starts with a chatbot. The co-founder of Imagine Entertainment — the company behind “A Beautiful Mind,” “Apollo 13” and “Liar Liar” — said he sits down with Anthropic’s AI assistant, Claude, to rough out a story before handing it to a writer.

“You can build the whole thing into an outline. You still need a screenwriter. I always believe you need a screenwriter,” Grazer said during a keynote at UCLA’s Entertainment Symposium on Thursday. What once could have taken up to a year, he said, now takes him about a week — but the human writer stays.

That balance — AI as an accelerant rather than a replacement — captures where much of Hollywood has landed in practice. Amazon MGM, Lionsgate, Netflix and Disney have all made major investments in the technology. The sharper question at the symposium, which drew many of the industry’s top lawyers and dealmakers to the Westwood campus, was not whether to use AI but how: who authorizes it, how far it goes and who gets paid.

For the companies building the tools, the answer increasingly comes from the client. Studios, production companies and distributors regularly approach Promise, a generative AI company, to bring AI into their productions, and each arrives with its own usage guidelines, said the company’s president, Jamie Byrne. Those rules govern which AI models Promise may use and what protections apply — effectively letting each client decide how heavily AI figures into the work.

“It comes down to a risk appetite,” Byrne said during a panel on AI. “We know that there’s talent that are staunchly against it. We know that there are many who are okay with it.”

He framed adoption as a competitive necessity: “Every time there’s a technology change, certain studios or production companies rise. Others fall, and it’s usually the ones that are not leaning into the new tool.”

Ron Howard, also of Imagine Entertainment, argued the limits will ultimately be set elsewhere — by viewers. “Sure, it’s about efficiencies and budgets, but more than anything, audiences are going to tell us where those restrictions are,” he said. He expects AI-generated content to settle into its own subgenre over time, with audiences signaling what they will accept.

The most contested ground is labor, where consent has become the dividing line. The emergence of synthetic performers such as Tilly Norwood has made AI a central issue in SAG-AFTRA’s contract. The union’s most recent agreement draws a clear line between authorized digital replicas, which use a performer’s likeness with their consent, and fully synthetic creations.

Talent agencies are organizing around the same principle. In recent years, Creative Artists Agency began digitally scanning clients into what it calls the CAA Vault, building a replica of a client’s image, likeness and voice while leaving the talent in complete control of how it is used.

That control is beginning to carry real value, said Tammy Brandt, CAA’s deputy general counsel, who said she is seeing more deals that involve digital likeness. Hollywood has been slow to work out how to monetize these replicas, she said, but once it does, audiences will start to encounter them more often.

“You have to lean into the technology and understand what it can do, and honestly, how you can make money, work with talent and with creative assets in a way that the user is interested in,” Brandt said. “There’s a little bit of trial and error as you go with that.”