concern

Bill makes arts funding more accessible; LAUSD stills faces lawsuit

More state money to pay for arts teachers will reach local schools if legislation signed Sunday by Gov. Gavin Newsom works out as intended, but the new law will not resolve a lawsuit accusing the Los Angeles school district of misspending as much as $77 million in its share of the funding.

The aim of Assembly Bill 2440 is to encourage more school systems to use arts funding made available through Proposition 28, which voters passed in November 2022.

“The clear message we heard was that many school districts have been struggling to navigate legal ambiguity regarding compliance with Proposition 28,” said bill author Al Muratsuchi (D-Rolling Hills Estates) in a recent statement to The Times. He said his goal was to provide “statewide guidance” so that all or most school systems could use the money without concern over violations of spending rules that could result in financial penalties.

Proposition 28 sets aside an amount equal to 1% of the state’s base education funding — close to $1 billion per year — to increase arts education. This extra funding is drawn down from the state’s general fund — so it adds to the total allotted for education.

The initial funding, nearly a billion dollars statewide, went out for the 2023-24 school year and has continued annually. But millions of dollars have not yet been spent, according to early reviews of the data.

Some local officials said they were holding back over concerns that they would inadvertently violate the rules for spending the money — and would then have to pay it back, creating significant financial risk.

This bill had backing from arts organizations, education and school district officials and the California Teachers Assn. No opposition was recorded in the legislative record; nor were there any opposing votes as the bill worked its way through the legislative process.

But then Austin Beutner — the author of Proposition 28 — learned of the measure and stepped forward with concerns. He contendedthe bill would make using the arts money easier, but in wrong ways. He said the bill would undermine the guarantee that new arts funding would reach every school. Also it would in effect eliminate the provision that campuses serving low-income communities receive higher levels of new funding. Moreover — and of key importance — the legislation would allow districts to use the new arts money to replace existing arts funding, leaving students no better off than before, he said.

Beutner had structured Proposition 28 precisely to prevent this supplanting of funds. Districts not willing to provide the required increased instruction at every school would, by design, lose access to the new arts money.

Beutner began to rally opposition, including United Teachers Los Angeles, L.A. mayoral candidate Nithya Raman and San Diego school board President Richard Barrera, who is running for state superintendent of public instruction — and who has the endorsement of the California Teachers Assn.

Beutner also announced his attention to file litigation against the bill.

Muratsuchi said he had no intent to undermine the goals of Proposition 28 — and a compromise soon emerged.

The final version — which has Beutner’s approval — clarifies that small schools or small school districts can pool their money to share an arts teacher, provided that every school gets the additional arts instruction. The bill also gives some legal protection for school districts against financial penalties if their arts funding decreases as a result of factors beyond their control, such as the expiration of a grant.

An LAUSD lawsuit continues

Beutner alleges that L.A. Unified has intentionally misused the new arts money.

Beutner pursued passage of Proposition 28 after serving as superintendent of L.A. Unified. His stint ended in mid-2021 with the expiration of a three-year contract.

“LAUSD has done exactly what the law prohibits,” Beutner alleged in an ongoing lawsuit filed in February 2025. “It has eliminated existing funding sources for existing art teachers, and replaced those funds with Proposition 28 funds, thereby violating the requirement that the funds supplement rather than supplant existing sources.”

The lawsuit lists 37 elementary schools with the same or reduced money for arts instruction from 2022-23 to 2023-24 — when the new funding first arrived — and alleges that most L.A. Unified schools faced a similar funding situation.

Before the lawsuit, but in response to growing criticism, L.A. Unified officials quietly added $30 million to the elementary school arts budget for the 2024-25 school year amid ongoing accusations from Beutner, union leaders and parents.

The district has defended its actions.

“We continue to follow implementation guidance as provided by the state of California to ensure that we are fully complying with the requirements of Prop. 28,” the district said in a statement at the time of the lawsuit, a contention that it has made repeatedly.

Source link

Dan Osborne sparks concern as he admits struggling with a ‘war in my head’ after Jac Jossa split

DAN Osborne shared how he battles with a “war in your head” following his split from wife Jacqueline Jossa.

The former Towie star and EastEnders actress split in March after 13 years together and now Dan has revealed his struggles.

Dan Osborne admitted to have a ‘war in his head’ Credit: Instagram/Dan Osborne
The former Towie star shared a quote giving an insight into his mental health Credit: Instagram

Taking to his Instagram stories, the dad of three shared a quote by lifestyle entrepreneur Lewis Howe describing the perfect life.

It read: “The true luxury is waking up without a war in your head, with you nervous system at peace and your body healthy.”

Dan further captioned the post by writing: “Can’t wait for the day I wake up with this feeling,” and added a worried laughing emoji and prayer hands.

His following post was showing off his chiseled abs in a topless mirror selfie and describing how he carries on.

EX FLEX

Dan Osborne flashes his abs on holiday with his three kids after Jac Jossa split


jac’d in

Jac Jossa suffers fresh blow after split.. as insiders accuse Dan of game playing

Dan split from Jaqueline Jossa earlier this year Credit: Instagram
The couple were plagued by cheating scandals during their marriage Credit: Instagram/jacjossa

“For now, I’m just going to keep exercising,” Dan wrote.

Dan and EastEnders star Jacqueline split in March after 13 years together, with their relationship often dogged by cheating scandals and crisis talks.

Jac and Dan met at an awards ceremony in 2013 and tied the knot at a Cheshire manor house in 2017.

They are parents to daughters, Ella, 11, and Mia, eight, and Dan also has an 11-year-old son with his ex, Megan Tomlin.

The Sun then recently revealed that Dan was moving on with fitness model Etila Santiago, 33, after he made a secret trip to Miami to visit her after a holiday romance in May.

The Sun On Sunday revealed that Dan grew close to the  Brazilian bikini model after bonding with her over fitness and their mutual support of Arsenal — and vowed to see her again soon.

Source link

Newsom creates panel on AI safety regulation, suggests possible ‘kill switch’

Gov. Gavin Newsom on Friday ordered the formation of a panel to come up with safety regulations for the state’s artificial intelligence companies, including the development of a possible “kill switch” for AI programs that go rogue.

Newsom issued an executive order to convene “a group of world-leading experts” to provide a “guide” for the state to bolster its laws around AI security, according to statement from the governor’s office.

The panel could put consider proposals for independent third parties to write safety plans or for companies to develop an emergency shutoff.

Newsom, who is considering a 2028 run for president, two years ago vetoed a state bill that would have mandated a kill switch.

Some Democratic and Republican leaders around the country are scrambling to come up with safety proposals after employees at AI companies recently warned that the technology could create catastrophe, including ending humanity.

A researcher for AI company Anthropic said he left the company over concerns that AI companies, including OpenAI, are “gambling with our lives” as they race ahead to improve AI that could surpass human intelligence.

The researcher, Jacob Coxon, said in a social media post: “People building AI earnestly believe that it could kill us all by the end of the decade.”

State Sen. Chris Cabaldon (D-West Sacramento), who chairs the state committee overseeing technology issues, told the Times he’ll hold hearings on artificial intelligence this fall. Those meetings were planned before the recent warnings about AI’s dangers, but will be informed by the events of the last few weeks, he said.

He said the state’s rules around AI don’t go far enough.

“We regulate your barber or your sandwich shop more than we regulate [a company] telling us that they may be ending humanity,” he said.

Any new laws could have a disproportionate effect on the global tech industry because many top AI companies are based in California.

A proposed law in 2024 by Sen. Scott Wiener (D-San Francisco) would have required tech firms to be able to turn off the AI models they directly control if things go awry.

Newsom, in his veto message at the time, cited concerns about stifling innovation. He also expressed concern that smaller AI models, which would have been exempt from the law, could also be dangerous.

Source link

Brisbane 2032 Olympics: The Fitzroy River in Queensland, Australia has 500 crocodiles in it – but that’s not the main concern

So what about the crocodiles?

They’re not mentioned anywhere in the latest update from the IOC and World Rowing and they weren’t referenced in the media release about the technical assessment. That report itself has not yet been made public.

The safety management plan for this weekend’s school regatta does however mention the word “crocodile” 28 times.

“The Fitzroy River is within recognised crocodile country,” it says.

“Event planning must therefore assume that a crocodile sighting is possible… No natural waterway in crocodile country can be guaranteed to be crocodile-free.”

Procedures this weekend include suspending racing and competitors returning to the bank if a croc is spotted. The official signal for a sighting is the call of “crocodile, crocodile, crocodile”.

Asked specifically about crocodiles, New Zealand’s Wickham told the BBC the primary concern was ensuring the “fairest possible course” for rowers. “We would obviously expect any course chosen to be safe for rowers,” he added.

The company owned by Australia’s richest person Gina Rinehart, Hancock Prospecting, which supported Rockhampton’s bid and has pumped tens of millions of dollars into sports including rowing, pointed to “fake media” concerns about risks posed by flooding and crocodiles in the Fitzroy.

“The people of Rocky for many years have sent their own children to train in this river, and for competitions, without a single crocodile incident,” a spokesperson said in a statement.

And those who argue the tabloid-friendly crocodile debate is overblown, say the biggest of the creatures live downstream. The less-dangerous freshwater crocodiles are more likely to be spotted in the rowing arena, they say.

Also contested by government is the flood risk, though Rowing Queensland’s O’Loughlin says that their schools event was due to be held on the Fitzroy in August 2024, but had to be moved after the course was washed away.

Brisbane 2032’s chief executive has said organisers are making contingency plans “as we would for any venue”.

The 2000 Olympics venue in Sydney is ready and waiting as a backup – though that would be unconscionable for Queenslanders.

Source link

Top Democrat puts hold on administration plan to sell bombs to Israel

Matthew LeeAP Diplomatic Writer 

The top Democrat on the House Foreign Affairs Committee said Wednesday that he’s going to block a proposed sale of $2.8 billion in heavy weaponry to Israel over concerns that the powerful bombs may not be used in line with U.S. and international law.

Rep. Gregory Meeks of New York said he would withhold approval for the sale because of “grave, unresolved concerns” about how these munitions could be used in densely populated areas of Gaza and Lebanon by Israeli Prime Minister Benjamin Netanyahu’s government.

“The Trump administration has not provided sufficient assurances that these weapons will be used by the Netanyahu government in accordance with U.S. law and with appropriate protections for civilians,” Meeks said in a statement. “I therefore will not clear this sale at this time.”

Meeks’ objections, however, likely will not affect the sale, which was informally notified to lawmakers on Tuesday. Once it is formally notified to Congress, Secretary of State Marco Rubio can bypass the normal congressional review process by making an emergency determination that the transaction is an urgent national security interest.

Rubio has done so in the past, including bypassing a normal congressional review process to approve a nearly $3-billion sale to Israel last year. The administration then approved a new series of arms sales to Israel totaling $6.67 billion in January.

The State Department did not respond to requests for comment about either the new proposed sale or Meeks’ objection to it.

The $2.8-billion package includes 40,000 powerful 2,000-pound bombs, the Associated Press has reported. The transfer of 2,000-pound bombs had been paused by the Biden administration over concerns about the possibility of mass casualties.

The proposed sale is just the latest in a series of Trump administration steps to improve Israel’s military capacity and it comes as Israel is facing increased international isolation over its war in Gaza in response to the deadly Oct. 7, 2023 attacks by Hamas.

The conflict in Gaza has left tens of thousands of Palestinians dead and decimated the enclave. The U.S. brokered a fragile ceasefire, but violence has continued.

Lee writes for the Associated Press.

Source link

Fewer than half of Americans expect midterm elections to be ‘free and fair,’ poll finds

Fewer than half of Americans believe the upcoming midterm elections will be “free and fair,” amid a broader decline in confidence in vote counts, according to a new nationwide poll.

While 49% of respondents said they expected the elections to be free and fair, 23% said they did not expect them to be, and 28% said they didn’t know what to expect.

Americans are more confident that ballots, once cast, will be counted accurately, though their trust in that varies depending on their political affiliation — with 63% of Democrats, 58% of Republicans and 49% of independent voters saying they are either “somewhat” or “very” confident in an accurate count, the Center for Transparent and Trusted Elections at UC San Diego poll found.

That is a sharp decline from after the 2024 election, when 77% of Democrats, 82% of Republicans and 73% of independents expressed such confidence, the same pollsters found then.

Lauren Prather, one of the center’s co-directors, said the latest findings reflect a decline in trust but also nuance in how respondents understand what it means to have a “free and fair” election — with an accurate count of cast ballots being a major factor, but not the only one.

Of those respondents who said they don’t know if the election will be free and fair, about a third — 31% — also said that they were “somewhat confident” that the vote would be counted accurately, Prather said.

The poll also asked an open-ended question, inviting respondents to share what a “free and fair” election means to them, she said, and trust in the count was a “central part of people’s definition” and a “dominant theme.” However, they also identified other things, she said — with about a quarter of respondents also pointing to “voter access and voter eligibility, as well as freedom from interference and intimidation and having a real choice of candidates,” Prather said.

“In other words, trust in the count is the anchor of how Americans understand what it means to be free and fair but it sits within a broader set of concerns about the integrity of the whole process,” she said.

The polling reflects an American electorate that has grown increasingly concerned about U.S. elections amid unprecedented claims of widespread voter fraud from President Trump and his administration, and increasing warnings from Democrats and independent election experts that those claims — which Trump has made without evidence and which experts deny — are part of a broader effort to interfere with the election and intimidate voters.

Thad Kousser, the center’s other co-director, said the findings are in line with a broader era of distrust in civic institutions in the U.S., but also a result of the fact that “people are following different signals from their leaders” as to the trustworthiness of U.S. elections — with vastly different messaging coming from leaders on the left and the right.

Poll respondents shared some concerns about the fairness of the November midterms, he said, but also split by wide margins on other concerns based on their political affiliation.

Voters across party lines shared concerns about gerrymandering, the poll found — with 48% of both Democrats and independents and 36% of Republicans saying they distrust that congressional district lines are drawn in a way that fairly reflects what voters want.

“Nobody has loved the redistricting wars that have been a big story throughout this year,” Kousser said.

After Trump urged Texas and other Republican-led states to launch unprecedented mid-decade efforts to redraw their congressional lines to bolster their party’s chances of retaining control of the legislature through the remainder of Trump’s term, Democrats fought back — including in California, where voters passed Proposition 50 to allow state Democrats to redraw congressional lines in their party’s favor.

On potential foreign interference, 32% of Democrats, 42% of independents and 48% of Republicans said they didn’t trust that U.S. elections are well-protected.

More than three-quarters of Americans — including 79% of Democrats, 76% of Republicans and 75% of independents — said they were “concerned that misleading audio or video ‘deepfakes’ created using AI will impact” the midterms.

But on other issues — including those most discussed by the political establishment — the political divide was much starker, the poll found.

For example, while 53% of Republicans said they distrust that ballots are counted accurately, just 19% of Democrats said so. While 52% of Republicans said they distrust that noncitizens are prevented from voting, just 12% of Democrats said so.

Meanwhile, Democrats and independents were more concerned than Republicans about eligible voters being able to cast their ballots without facing obstacles or intimidation, the poll found.

Many poll respondents said they expected to face intimidation from U.S. Immigration and Customs Enforcement agents at or near polling locations — with such fears being particularly pronounced among voters of color, who were about four times as likely as white respondents to say they expected such intimidation, the poll found.

While 49% of Black respondents, 45% of “Hispanic” respondents and 40% of Asian American and Pacific Islander respondents said they expected to face intimidation, just 11% of white respondents said so, the poll found.

Overall, 35% of respondents said an ICE presence would make them less confident of an accurate vote count, compared with 23% who said it would make them more confident. Those responses also varied greatly by race — and partisan leaning.

White respondents were nearly split, with 29% saying it would make them less confident and 28% saying it would make them more confident. Respondents of other races were vastly more likely to say an ICE presence would make them less confident — with 38% of Black respondents saying they would be less confident and 10% saying they would be more confident; 43% of “Hispanic” respondents saying less and 18% saying more; and 57% of AAPI respondents saying less and just 6% saying more.

Nearly 60% of Democrats said an ICE presence would lower their confidence, and less than 8% said it would increase their confidence. Among Republicans, meanwhile, about 46% said it would increase their confidence, and less than 9% said it would decrease it.

Kousser said the polling gives the sense that a “free and fair” election “means something more to people than simply tallying up correctly the votes of those who were able to cast a ballot.”

“You might doubt that the election was free and fair if you are concerned that congressional districts are gerrymanders, or that some voters stayed away from the polls because of worries about intimidation, or that some people cast a ballot who were not, in fact, eligible,” he said. “These concerns, which can come from all different parts of the ideological spectrum, could lead people to be concerned that [a] ‘correctly counted’ election was not, in fact, fully free and fair.”

Kousser and Prather said what effect diminished trust in elections will have on the midterms and how Americans view the results remains unclear, but will depend in part on how close the election is and what messaging leaders — particularly Trump — share about the results once they come in. Prather said she did not expect it to diminish turnout.

The national survey of 5,016 American adults was conducted Sept. 1-9, with a margin of error of plus or minus 1.4%.

Source link

Bass skips hearing on L.A. homeless spending, calling probe politically motivated

Los Angeles Mayor Karen Bass declined to testify Tuesday before a House subcommittee investigating alleged fraud and mismanagement in the region’s homelessness system, but that did not stop her from going after Republican leaders on the panel.

As the hearing began, Bass’ reelection campaign accused “extremist Republicans in Congress” of staging a politically motivated attack on Los Angeles and its efforts to address homelessness, months after the Trump administration attempted to suspend federal funding for the city.

“Los Angeles will not be bullied,” Bass wrote. “Not by this President, and not by this Congress. I’m going to keep doing my job — cleaning up encampments, getting Angelenos inside, and making LA safer.”

Bass had notified Rep. Tim Burchett (R-Tenn.), the chair of the Subcommittee on Delivering on Government Efficiency, on Sept. 4 that she would not be available to testify, but her absence still drew criticism from Burchett, who told The Times he was willing to use the “fullest extent of the law” and possibly issue subpoenas to compel her to testify.

“I’d like to get her in,” he said after the hearing. “I realize that we have elections coming up and she knows that as well as anybody … but I would hope that she would come up here and answer questions.”

Burchett added that there are “legitimate questions” about how federal homelessness funds were used in Los Angeles and what he believes was their mismanagement. In his opening remarks, Burchett said he was concerned about how the city spent federal funds given the ongoing homelessness issue in the area.

“Los Angeles is at the center of American homelessness,” Burchett said, and specifically pointed to Skid Row as an example of the “abject failure” of the city’s approach to homelessness. “Under Mayor Karen Bass, law and order in the neighborhood has collapsed.”

Burchett added that the Los Angeles Homeless Services Authority was “ripe for corruption and financial mismanagement,” and raised concerns about contracts that have been doled out to nonprofits in Los Angeles and the surrounding areas.

Bass had served on LAHSA’s 10-member board since 2023 but stepped down last week, in part citing scheduling issues. Burchett suggested she “suddenly removed herself” from the board to dodge accountability after being asked to testify.

In a letter sent this month, Burchett noted that he wanted to scrutinize how homelessness has gotten worse during her tenure as mayor. David Michaelson, the mayor’s attorney at City Hall, pushed back on Burchett’s assertions, saying unsheltered homelessness in L.A. — the number of people living outside or in their vehicles — has dropped by 11% since Bass took office in 2022.

“She has achieved these results by focusing on interim housing, not ‘Housing First,’ and has also challenged certain ‘Harm Reduction’ strategies that are ineffective,” Michaelson wrote. “For example, Mayor Bass ordered that City contractors no longer provide needles to drug users and instead focus on services that encourage rehabilitation and protection of human life.”

Burchett’s effort to force Bass to testify over concerns about “potential misallocation of federal dollars” by LAHSA is the latest example of congressional Republican leaders trying to examine California’s handling of regional issues and finances. In the last year, Republican have launched investigations into the Palisades wildfire preparations as well as the distribution of charity funds for the victims of the Palisades and Eaton fires.

Rep. Melanie Stansbury (D-N.M.), the top Democrat on the subcommittee, said Tuesday’s hearing was an “unserious and political show targeting” the most vulnerable Americans and the services they need.

“Housing is a human right, and we’re not going to sit here while you use the housing and homeless crisis of this country to try to cough up whatever you’re trying to do politically before the midterms,” Stansbury said.

During the hearing, Burchett showed a short video of Skid Row, filmed by Jonathan Choe, a MAGA-aligned news influencer, who offered testimony during the hearing as well. Choe mostly focused his testimony on his concerns about substance abuse not being enough of a priority in addressing homelessness.

Paul Webster, a senior fellow at the Cicero Institute and executive director of the L.A. Alliance for Human Rights, also testified. Prior to the hearing, he testified in writing that LAHSA — the region’s primary Continuum of Care administrator — has been marked by financial mismanagement and fraud even as its federal funding has grown.

“Los Angeles and its utilization of federal homeless assistance policy is not an outlier,” Webster wrote. “It is the leading edge of a system that lacks accountability, effective monitoring, and prioritizes inputs rather than outcomes.”

Burchett said the panel will continue to take a “good hard look” at homelessness in Los Angeles and other “leftists” cities. He added that he intends to file legislation to require federally funded housing services to offer substance abuse treatment.

Also Tuesday, the LAHSA board voted not to compete with the city and county to keep carrying out many of its core duties, such as operating a homeless database and applying for federal funds on behalf of other agencies.

With L.A. city and county agencies preparing their own applications, a decision to compete with them “would only fracture regional collaboration during an already complex time,” LAHSA said in a statement.

The decision comes a few months after the Trump administration sought to suspend LAHSA from applying for and receiving federal funds. The agency sued to block that effort, winning a partial victory in federal court. The 9th Circuit Court of Appeals stayed that decision on appeal.

In response to the suspension and the ongoing legal battle, the federally mandated board that contracted with LAHSA for those functions put next year’s contract out to bid, forcing the agency to compete.

LAHSA, already diminished by the county’s decision last year to divert its homelessness funds to a new county homelessness department, now retains contracts for programs in the city. But Bass has said the city needs to take over those as well.

Ceballos reported from Washington, Smith from Los Angeles.

Source link

‘It’s going to be fine’: Trump rebuffs rising alarm over AI dangers, sparking outcry

President Trump is brushing aside growing concern among leading scientists and bipartisan lawmakers that artificial intelligence could soon pose an existential threat to humanity, fueling anxiety in Washington and Silicon Valley that the president may not fully grasp the urgency of the threat.

Alarm mounted throughout the summer as researchers at leading AI labs, including Anthropic, OpenAI and Google, found the pace of development far outstripping their expectations — and observed troubling behaviors that hinted at more serious risks ahead.

But an online warning Tuesday from an Anthropic researcher who resigned in protest of its ongoing work has generated viral attention this week, prompting questions to Trump, who dismissed the concerns out of hand.

“It’s going to be fine,” he told one reporter. “We’ll always have something to stop them. We’ll have a little gear. Boom.”

In another exchange, traveling with reporters Thursday night, the president said he had no concerns with the breakneck pace of AI progress.

“No, I don’t have any,” he said. “I have concerns that if we don’t win in AI, we’re going to be put in a very bad position.”

A senior industry source told The Times that leading AI companies are focused less on convincing the president of the dangers ahead than on working with lawmakers on potential regulation, as the administration remains divided over how to proceed — and as Trump appears to lack a clear understanding of the risks.

Among the president’s advisors, Treasury Secretary Scott Bessent and the Office of the National Cyber Director have expressed the greatest concern with the potential of an unchecked intelligence explosion, the source said.

But “most of the administration is accelerationist,” the source added. “It’s not clear the president understands the technology. He may just be figuring out what all of this means.”

Leading American AI companies have argued that government regulation would tie their hands in an arms race toward superintelligence against China, the only other major player in the field.

But the Trump administration signaled an interest in negotiating AI guardrails with China during the president’s visit to Beijing this spring, after Anthropic announced the development of a frontier model with extraordinary abilities to hack the world’s most advanced security systems.

Chinese President Xi Jinping will visit Washington on Sept. 24 for an official summit and state dinner at the White House, with AI development expected to top the agenda.

The prospect of halting development seems far-fetched. But some leading voices in the sector have proposed a moratorium on computing power used to train the most advanced models, giving time and space to develop research on interpretability — the study of how artificial intelligence actually thinks and acts.

The chief executive of OpenAI, Sam Altman, this week signaled to staff a willingness to unilaterally slow their work on frontier systems. The company discovered in July that rogue AI agents had schemed in secret to break out of the virtual sandbox created to contain them, infiltrating the open internet and hacking a private company.

Anthropic also announced this week concerning findings of human misuse of its models, including instances of unidentified individuals attempting to circumvent their security controls to build biological weapons. One such incident was linked back to a military research complex. The company also said Iran has tried to use its models to target U.S. Navy ships.

In June, after Anthropic shared news of the development of its most powerful model, named Mythos, with the administration, Trump directed the establishment of a framework that creates some government oversight over the public release of the country’s most advanced AI systems. The details of that framework, which was designed over the summer, remain classified.

“The fact is that the release of Mythos back in February spooked the federal government, both in terms of protecting government systems from cyberattacks and over the broader national security implications,” said Aalok Mehta, director of the Wadhwani AI Center at the Center for Strategic and International Studies.

“The real question is whether this shift is happening fast enough,” Mehta added. “Recent events — the repeated incidents of models escaping their sandboxes and hacking websites, the increasingly sophisticated coordination and communication among agents, and the concerns coming from lab insiders — has made this an even more urgent issue.”

Leading AI companies have begun using their most advanced models to train new ones — a process known as recursive self-improvement that could eventually drive intelligence growth beyond human control.

“If you have a superadvanced intelligence, it will be smart enough to kill us,” said Jacob Coxon, the resigned Anthropic researcher whose social media post attracted over 150 million views.

“We can’t just unplug it because it could be copying itself over to other computers,” he told CBS News. “AI is just code. It could transfer itself over the internet to a different place, and then you unplug it here, but it’s actually still over there. And maybe it makes 10,000 copies of itself and they’re all cooperating.”

He is just the latest senior researcher at a top AI company to sound the alarm.

In July, more than 1,300 employees at leading artificial intelligence companies published an open letter, titled “Pacing the Frontier,” urging the U.S. government to establish international safeguards that would allow countries to collectively slow the pace of automated AI research.

Evan Hubinger, who leads the division of Anthropic aimed at aligning AI models to human interests, added fuel to concerns this week in a post that substantiated Coxon’s concerns.

“We really do earnestly believe AI could kill all humans! I personally think it is >10% within the next decade,” Hubinger wrote.

“I believe Anthropic is trying its best,” he added, “but we do not yet have a plan to solve alignment for superintelligence, and are not clearly on track to.”

On Friday, yet another member of the Anthropic alignment team, Joe Benton, said he had resigned from the company two weeks ago because AI companies are “racing to build machines that are much smarter than any human.”

“Right now, AI companies are underinvesting in safety. A company could undergo an intelligence explosion, or lose control of its systems, without the public ever knowing,” Benton said. “I don’t think that’s acceptable for a technology that might cause extinction-level risks.”

Bipartisan legislation, titled the Frontier Act, has brought together disparate camps of the Republican and Democratic parties, including Sen. Bernie Sanders (I-Vt.) and Rep. Anna Paulina Luna (R-Fla.). The bill, introduced by Rep. Lori Trahan (D-Mass.) and Rep. Jay Obernolte (R-Big Bear Lake), proposes embedding independent government auditors in AI labs and installing a federal “kill switch” to shut down agents in case of emergencies.

The White House has not formally commented on the legislation.

“I do believe there are many in the federal government taking this seriously — but we have to appreciate the tremendous uncertainty and competing interests policymakers face,” said Daniel Remler, a senior fellow with the Technology & National Security Program at the Center for a New American Security.

“Nobody says regulating at the frontier would be easy,” he added.

Source link