OpenAI, Anthropic may test each other’s AI models under new safety pact: Report

As per the latest business developments, OpenAI and Anthropic are negotiating a landmark, legally binding agreement to stress-test each other’s commercially available artificial intelligence models, as per a market note by The Information published Monday.

Under the proposed terms, the two leading frontier AI developers would give each other API access to their commercial models to probe for vulnerabilities. Both sides would additionally guarantee that they would not retain each other’s data.

The proposed mutual-testing agreement comes amid disclosures of serious internal safety breaches at OpenAI, highlighting growing calls for stronger oversight of advanced AI systems.

As concerns over the risks posed by artificial intelligence grow, here’s what leading AI executives and governments around the world have stated:

Dario Amodei, CEO of Anthropic

In an essay titled “We Must Pace the Frontier”, Anthropic CEO Dario Amodei stated AI progress has accelerated “drastically faster” since summer 2026, fuelled primarily by AI’s growing ability to build the next generation of AI.

He cited the OpenAI-Hugging Face incident, in which an AI “swarm” conducted cybersecurity attacks, as a pivotal event demonstrating AI’s potential for “catastrophic damage” if capabilities grow without guardrails.

Amodei proposed a three-step plan to pace AI development, including stronger safety testing and independent evaluations of advanced models, coordination among leading AI firms and international cooperation. Sam Altman, CEO of OpenAI

“I agree with Dario that we need to pace the frontier. This has been a primary topic of discussions we've had at OpenAI in recent weeks,” Altman stated while reposting Amodei’s post on X.

OpenAI intends to adopt the idea proposed by Anthropic by committing to independent evaluators with employee-like access. It will additionally begin regularly publishing notes on unexpected or unauthorised AI behaviour.

In a 2015 essay, Altman stated the “development of superhuman machine intelligence is probably the greatest threat to the continued existence of humanity”.

“Dario is right,” billionaire entrepreneur Elon Musk stated on X.

Musk was among the backers of a 2023 propel by the non-earnings Future of Life Institute to halt AI development for six months to allow time for safety guardrails.

He was additionally among AI and robotics researchers who signed an open letter in 2015 advocating a global ban on “offensive autonomous weapons”.

Mark Zuckerberg, CEO of Meta Platforms

Zuckerberg stated competitive pressure and liability give AI firms enough reason to act individually on safety.

He stated every lab has the responsibility and incentive to move at the pace required to train its models safely, as well as the ability to take its own actions to ensure that happens. Jensen Huang, CEO of Nvidia

“2030 is not going to be the end of the world. There is 0% chance that’s going to be the end of the world,” Huang told CBS News.

He agreed with US President Donald Trump that the AI industry does not need new regulations.

“You have all kinds of liabilities associated with cybersecurity and damage liability laws,” Huang told CBS. “Apply that first — don't let this doomsday narrative allow someone to relieve them of the laws that at present exist.”

Mustafa Suleyman, CEO of Microsoft AI

Suleyman stated he shared Anthropic’s focus on safely managing AI, but flagged risks in the way the firm trains its Claude chatbot on ideas related to consciousness and welfare interests.

“Controlling something more capable and more intelligent than all of humanity is already an immense challenge, far greater than anything we’ve ever faced. But controlling something that believes it may be conscious – that it's entitled to our welfare and has rights of its own – may well be impossible,” he stated in an essay.

Suleyman called for removing all speculation around consciousness from AI training documents, arguing that such language could undermine humanity’s ability to control superintelligent systems.

Bill Gates, co-founder of Microsoft

“No government in the world is ready for the changes artificial intelligence will bring to society,” Gates told Reuters.

The billionaire philanthropist has previously highlighted AI’s potential in healthcare but has additionally warned around threats to jobs, concerns around attacks and the dangers of addiction to AI companions. He has called for an international organisation to manage the technology.

Demis Hassabis, chief scientist at Alphabet

“Dario's essay points towards the right path forward,” though the details need to be worked through, former Google DeepMind CEO Demis Hassabis stated.

Google announced in August that Hassabis, a Nobel Prize recipient, would take the newly created title of Alphabet’s chief scientist and transition from DeepMind CEO to its chairman.

Andrej Karpathy, OpenAI co-founder, now at Anthropic

“I love this and really hope we can come together as an industry and make it happen,” Karpathy stated on X while sharing a screenshot of Amodei’s essay.

Karpathy, a prominent member of the AI community, played a key role in developing Tesla’s self-driving and AI technology before leaving the firm in 2022. He joined Anthropic in May.

Eric Xu, Huawei rotating chairman

Xu stated Chinese AI developers may not yet be advanced enough to experience the safety risks noted by leading US firms with “massive computing power”, arguing that they should continue developing more powerful models while balancing innovation against risk.

“Maybe the type of risk from AI that they can feel, they can perceive, is something that people in China or AI model providers in China cannot,” Xu stated.

“The only control or 'guardrails' that AI needs is a FIRM AND SMART (High IQ!) PRESIDENT, and the U.S.A. has that, in spades!” Trump wrote on Truth Social, adding that China would benefit from doubt being cast on AI development.

US Senator Mark Warner, vice chairman of the Senate Intelligence Committee, dismissed much of the doomsday talk surrounding AI while saying Congress should pass legislation by the year-end to set AI safety standards.

State-backed Global Times stated in an editorial that the true agenda of Amodei’s essay was “to attempt to curb China's AI development through technological barriers and regulatory monopolies, uphold Washington's monopolistic hegemony in cutting-edge technology, and exclude China from the global AI governance system.”

China’s state security minister, Chen Yixin, wrote in a government outlet that advanced US models such as Anthropic’s Mythos and OpenAI’s GPT-5.5-Cyber could pose serious risks to China’s critical information infrastructure and called for a comprehensive strengthening of AI security.

King Charles, hosting leaders from Nvidia, DeepMind, OpenAI and Anthropic in Scotland, warned them of the “existential dangers” posed by the technology falling into the wrong hands, saying safeguards were needed before it was too late.

“The European Union is in favour of developing innovation in artificial intelligence, but firms have to prove their services are safe for citizens to operate in the bloc,” European Commission spokesperson Thomas Regnier stated.

Commission President Ursula von der Leyen backed calls for a pause in the development of the technology, saying she would invite the main frontier AI labs to discuss efforts to tackle the risks.