platform

RIYADH: TikTok has taken down 15,519,240 videos across Egypt, Saudi Arabia, the UAE, Oman, Iraq, Morocco, Lebanon and Sudan during the first quarter for Community Guidelines violations.

The move reflects the scale of the platform’s enforcement efforts and its focus on maintaining safety and accountability across the Middle East and North Africa region, a statement said.

As part of its broader efforts to promote age-appropriate and “authentic experiences” on the platform, TikTok removed 86,288,705 fake accounts and 25,764,372 accounts suspected of belonging to users under the age of 13 globally during the first quarter of 2026.

“These actions reflect the scale of TikTok’s efforts to identify and remove accounts that may not meet its platform requirements, while strengthening safeguards for younger members of its community,” TikTok said.

“Through a combination of automated detection, account-level enforcement, and human review, TikTok continues to reinforce the systems designed to support a safer and more age-appropriate experience.”

Globally, the platform suspended more than 50 million LIVE sessions from January to March, an 18.6 percent increase from the previous quarter. A further 58 million sessions were subject to demonetization or warnings, while 21,966,667 LIVE creators were also demonetized or warned.

Across the eight MENA markets, TikTok said it proactively interrupted millions of livestreams and banned hundreds of thousands of LIVE hosts for violating its guidelines.

In Saudi Arabia, TikTok removed 2,811,302 videos in Q1 2026, with a proactive removal rate of 98.2 percent and 92.8 percent of violative content taken down within 24 hours of being reported. The platform restored 150,916 videos following successful appeals, and also banned 48,106 LIVE hosts while interrupting 371,758 livestreams.

Iraq recorded the highest removal volume in the region. TikTok removed 5,253,566 videos that violated its Community Guidelines during the quarter, with a proactive removal rate of 99.4 percent and 95 percent of violative videos taken down within 24 hours of being reported.

California lawmakers have passed legislation that will make it easier for government agencies to protect themselves from the rising use of artificial intelligence for public comment, records requests and other forms of civic engagement.

Senate Bill 1159 from Sen. Christopher Cabaldon (D-West Sacramento) prohibits anyone from knowingly using AI to falsely represent that a real person engaged with a government agency. It also specifies that agencies are not required to treat engagement from AI or bots as if they were real humans.

“What we have seen with the result of the advent of artificial intelligence and other similar technologies is the capability for these systems to flood the zone — to drown local governments, and potentially state agencies as well, in inauthentic, non-human engagement” Cabaldon said during a March meeting of the Senate Judiciary Committee.

The legislation was introduced shortly after a February report from The Times about a campaign to sway a vote on gas-powered appliances at the South Coast Air Quality Management District. A Southern California based public affairs consultant named Matt Klink took credit for the campaign, stating that he used a platform called CiviClick to flood the district with 20,000 public comments opposing the rule ahead of the air board’s vote.

CiviClick describes itself on its website as “the first and best AI-powered grassroots advocacy platform.” Company officials maintain that AI was not used in the AQMD campaign, but said it is a tool they offer and use in other campaigns. Chief executive Chazz Clevinger said he could not share how the 20,000 comments to the air board were generated or how constituents were identified and contacted.

Agency insiders said the onslaught of emails almost certainly influenced the air board’s decision to reject the proposed rules, which would have imposed fees on new gas-powered furnaces and water heaters for some 10 million appliances across the South Coast region.

Cabaldon cited The Times story when he introduced the legislation, noting that at least three people contacted by the air district said they had not submitted the public comments attached to their names.

He also cited a report from the San Francisco Chronicle about a similar campaign to sway a different rule at the Bay Area Management District, which was run through a platform called Speak4 that advertises its ability to produce custom AI-powered letters.

The business advocacy group that ran the campaign also denied that AI was used. However, 10 people contacted by The Chronicle said they had not written the letters attributed to them. “This was forged,” one person said.

Reached by phone, Cabaldon said the legislation will help public agencies navigate how to respond to the deployment of AI, which is increasingly being used in a way that “swamps our civic engagement process, but also disables our state and local governments altogether.”

For example, the California Public Records Act requires government agencies to respond to requests for public records within 10 days, while the Brown Act and the Bagley-Keene Open Meeting Act guarantee the right to participate in public meetings and provide public comment.

“The point of the bill is to say that these laws are about humans, and just because it comes in the form that a human would write it, does not mean you have to treat each of these communications as if it’s a human being, and therefore, AI is not entitled to 10 days, AI is not entitled to three minutes at the school board meeting,” he said.

Experts said the use of AI for “astroturfing,” or faking, civic engagement is a growing trend. In the United Kingdom, a service called Objector.ai is using AI to identify and generate formal objections to local planning applications, garnering the concern of experts, The Guardian reported.

Public officials in California are worried, too. Vacaville vice mayor Michael Silva said the city has been receiving AI-generated public records requests, which are slowing its ability to respond to other legitimate requests submitted by residents.

Dylan Plummer, deputy director of the Sierra Club’s Clean Heat Campaign, said many AI campaigns have benefited the fossil fuel industry and pose an “existential threat to public participation in our democracy.”

“The passage of Senate Bill 1159 is an important step to clarify the law and discourage the use of emerging technologies to falsify public records and mislead regulators in California,” he said. “That said, much work still needs to be done to understand how widespread this practice is, and to hold bad actors accountable for laws that may have already been broken.”

Lawmakers acknowledged that the legislation is just a start, and that it is increasingly difficult for public officials to detect bespoke letters, deepfake videos or other kinds of engagement powered by AI. The bill authorizes government agencies to use disclosure verification tools to determine if AI is present — something the Bay Area Air District already indicated it may do by replacing its email system with a website for public comment submissions instead.

The legislation does not preclude real people from using AI to facilitate genuine public engagement, such as someone using ChatGPT or Perplexity to improve the text of a letter, so long as the volume and frequency of their engagement are consistent with ordinary participation from a real person.

For its part, CiviClick notes on its website that it supports SB 1159, and said its platform already complies with what the bill proposes.

Some lawmakers said the use of AI in a civic capacity represents a new frontier.

“If I’d have read this bill back when I was on the Sacramento County Board of Supervisors, I would have wondered what you were smoking,” Sen. Roger Niello (R-Fair Oaks) said during the March meeting of the Judiciary Committee.

“But that’s how things have progressed, and the development of technology will always outpace the development of defenses against the undesirable effects of technology,” he said.

SB 1159 passed the legislature this month and will head to Gov. Gavin Newsom’s desk for signature in September.

Meta settles $16.68bn lawsuit over child addiction claims, agreeing to major changes in Facebook and Instagram features.

Meta Platforms has agreed to settle a lawsuit that accused the company of designing Facebook and Instagram in a way that addicted children, misled consumers about safety, and collected personal data of children on the platform.

On Wednesday, the social media giant agreed to pay a maximum of $16.68bn as part of a settlement to resolve claims brought in the United States case, championed by a coalition of 29 US states. The case, which started on August 18, was expected to last six weeks.

Meta, based in Silicon Valley in California, has also agreed to make changes to Facebook and Instagram nationwide as part of the settlement. Among these are daily usage limits of two hours for those under the age of 18, which can only be removed by a parent, and nighttime blocks.

The California State Attorney General’s Office said that the Mark Zuckerberg-led company would also identify and remove children under the age of 13 from the platform.

Meta denied any wrongdoing as part of the settlement, which still needs court approval. It had faced up to $1.4 trillion in fines in the case, but the coalition had been seeking a penalty closer to $200bn.

The settlement comes after a loss in a comparable landmark case in New Mexico, where a jury ordered Meta to pay $375m in March and another $567m in August.

Meta’s stock tumbled in early trading on Wall Street, down 0.1 percent since the market opened.

Things people don’t want to do this summer, as evidenced by poor ratings: Watch CBS news anchor Tony Doukopil. Tune into Paramount+’s sci-fi teen drama “Star Trek: Starfleet Academy.” Read President Trump’s Truth Social posts.

No matter how much Trump posts, and he has been posting a lot lately, traffic to the platform he uses as his megaphone for official White House statements and personal rants has fallen off significantly this summer. Last month, the overall number of monthly visitors to Truth Social was down about 36% from where it was in 2025, according to the online tracking firm Similarweb. The numbers were similarly dismal in June.

But Trump’s slumping media fortunes may soon get an infusion of cash, or bitcoin, or whatever it takes to line his coffers before the jig is up.

Never one to leave a source of income untapped, the president has come up with yet another way to add to the $2.2 billion he made in just the first year of his second term. His majority-owned Trump Media & Technology Group earlier this month announced that it was rolling out a new service aimed at cashing in on the president’s every word.

Truth API is a subscription service that offers early access to posts from Trump and other notable users of the platform, for a price. It’s charging fees of up to $100,000 and month.

But there’s a hiccup in the president’s latest grift. On Wednesday, media organizations Freedom of the Press Foundation and The Intercept sued Trump, filing a complaint saying that providing quicker access to his posts to those who pay was “extraordinary, corrupt, and unconstitutional.”

Their suit alleges that Truth API contradicts the First Amendment’s guarantee of equal public access to the president’s statements and violates the Fifth Amendment by granting preferential access for “unreasonable sums.” The lawsuit filed in the U.S. District Court for the Southern District of New York, asked the court to block Trump from publishing official government information exclusively on Truth Social.

So why is this particular money-making scheme garnering so much attention outside Trump’s many other grifts? Because a president’s words can, and often do, sway the stock market. In the frenetic world of Wall Street trading, early access to statements and news from the Commander-in-Chief gives subscribers an edge, and as NPR pointed out, that could mean a difference of millions of dollars.

Unlike any other sitting U.S. president, Trump in his second term has ignored traditional means of communication such as press briefings, live addresses or posting official announcements, executive actions, press releases, and statements on the official White House Website. He’s done so in favor of communicating through his own privately controlled platform, delivering wild posting sprees that often forgo the fact-based, informative briefings we the people still need from our elected officials. But even back when he was using Twitter (now X) during his first term, the White House said his tweets should be considered official statements.

That standard still holds for his frequent barrages of boasts, insults, threats, grouses and indecipherable dispatches via Truth Social. After the humiliating failure of his America 250 celebration, he fired off 67 posts on Truth Social in just two hours, posting almost every single minute between 11:12 a.m. and 1:14 p.m. His musings ranged from attacks on a federal judge to a photo of himself at a 1991 New York City tree-lighting ceremony with his “Home Alone 2” co-stars.

That spree is now among the thousands more posts from the president, that have not been followed up by announcements from the White House outside of Truth Social. “In other words, President Trump’s posts are the only way to get official government news,” the lawsuit said.

Trump Media & Technology Group, or TMTG, is majority-owned by the president. It was launched following Trump’s account suspension across mainstream social platforms including Twitter, Facebook and YouTube. The platforms cited risks of inciting violence following the Jan. 6, 2021, U.S. Capitol riot. Trump responded by creating his own platform, and Truth Social debuted in 2022.

But the platform’s parent company, TMTG, has lost money ever since it went public in 2024. On Monday, Trump Media reported a $238-million loss for the second quarter, tied mostly to cryptocurrency assets. Executives told investors on a conference call that they are now going to focus their energy on Truth Social and soft-explained their latest scheme to profit off the presidency.

“Our customers will get published and publicly available posts fractionally faster” than everyone else,” said Kevin McGurn, the company’s interim chief executive. He added that such early access is a “well-established business practice.”

Unless it’s a sitting U.S. president doing the selling. We’re in uncharted territory, once again. But another big question around this new subscription service is whether investors and traders can trust the intelligence they get from early access to Trump’s posts.

It was revealed this week that the president published deceptive information last month, putting the lives of dozens in danger. Before leaving a NATO meeting in Turkey, he posted that he’d be riding on the older Air Force One “for old time’s sake” instead of the newly retrofitted, Qatari-donated jet. His misdirection was part of an elaborate ruse to mask his transfer from Air Force One to a military fighter jet following intel that Iran may be targeting the president’s plane. The subterfuge involved him stowing away in an airport catering container to sneak onto the jet. Of course a president has to be protected, but Air Force One still had members of the press and his administration aboard when they sent it into the sky. Essentially, they were unwitting decoys.

Sometimes a president has to lie to stay safe. And often times this president peddles misinformation as a means to other ends, like amassing more money for himself while holding onto his seat of power. Paying for early access to Trump’s posts is a great idea — for Trump.

In its latest effort to protect human creators from the flood of synthetic music, Spotify will soon require AI-generated identities to carry an “AI Persona” label while barring their tracks from users’ algorithmic mixes.

The change will take effect in mid-September. It follows Spotify’s earlier moves to help subscribers understand the difference between genuine and artificial creators. The platform announced a verification badge in April designed to highlight human artists. In a statement Tuesday, Spotify said the new AI Persona label is meant to make the streamer “the most transparent and trustworthy place to listen to music.” Music from labeled AI Personas will also be excluded from the Swedish company’s personalized recommendations by default.

“While we believe all artists have creative choice in determining how they present themselves, Spotify’s programming is focused on elevating music from authentic artists building careers in music,” the company said .

The AI persona badge will soon appear on an artist’s profile, in the search feature and in playlists.

This is a step in the right direction, said Tiffany Naiman, the director of Music Industry Programs at UCLA. But, she said, she’s more interested in how the company will be identifying AI personas.

“I want to know the process. How [will they] know the difference? Are there going to be [real] artists that get tied up in it?” Naiman said. “It feels very ‘Blade Runner,’ right? Like you’re human, you’re not human.”

Starting Tuesday, Spotify users will be able to identify themselves as AI personas, if applicable. But the company said it “won’t rely on self-disclosure alone” and will use “human review alongside AI investigative tools to apply the label.”

Artists who get labeled as AI personas by Spotify will be allowed to appeal the label. In the coming months, users will also have the ability to report artist profiles as potential AI personas.

The label is the latest feature from the streaming platform that builds on transparency between the artist and their listeners. Recently, Spotify has also introduced SongDNA, an interactive feature that shows the creative team behind the track; AI Credits, a disclaimer where artists can reveal how much AI was used in their creative process and Artist Profile Protection, which allows artists to review and approve all releases on Spotify.

Artificial intelligence is becoming an essential part of Spotify’s business. The company announced a new AI deal last week with the digital music licensing company Merlin. The partnership will enable artists across the more than 30,000 labels and distributors in Merlin’s network to participate in Spotify’s upcoming AI tool. It hasn’t launched yet, but the goal is to let fans create AI-generated remixes and covers of existing songs on the platform.

Spotify is following in the steps of other streaming platforms that have taken more aggressive approaches to the new technology. Deezer, a French streamer, was the first to detect, tag and exclude AI-generated music from algorithmic recommendations. The company recently disclosed that up to 90,000 AI tracks are being uploaded to the platform daily, representing more than 50% of its new music uploads. Tidal has banned AI-generated music from receiving royalties on its platform.

“We know the music ecosystem is evolving, and so will our approach,” said Spotify in a statement. “We’ll continue to adapt as the landscape changes and as we learn from artists, listeners, and industry partners.”