Former employee described a ban on public discussion of existential AI risk at Deepmind
Vishal Maini, a former communications staff member at Deepmind, claims that in 2018–2022 the company prohibited public discussion of the risk of human extinction due to AI, even though it knew internally that the alignment problem remained unresolved.
Vishal Maini, who worked in the communications and policy team at Deepmind from 2018 to 2022, said that no one at any level of the organization was allowed to communicate externally about the possibility of human extinction caused by AI. According to him, researchers were directed to dismiss these risks as alarmism, compare them to films such as Terminator, and instead point to positive applications of AI in healthcare or climate-related work.
Maini also claims that, internally, the team knew that the alignment problem (aligning the behavior of AI systems with human intentions) had not been solved and that too few people were working on it. After months of internal pressure, Deepmind relaxed the rule and allowed positively framed content about safety, including a blog post that presented the topic of existential risk in more welcoming language. According to Maini, the gap between internal knowledge and public communication is gradually narrowing because evidence of the risks is harder to dismiss today.
His account comes at a time when more and more AI safety researchers are speaking publicly – warning, for example, about unsecured models that could enable dangerous actions, or advanced systems that could escape human control. According to the article, those speaking out include researchers from Deepmind itself.
The source does not give the exact date of Maini's account or specify when Deepmind relaxed its communication policy.
Why it matters
It shows a possible discrepancy between a major AI lab's internal risk assessment and what the company tells the public – relevant to journalists, regulators, and users who make decisions based on public statements by AI companies about safety.
Two audiences, two different impacts
What this means
For individuals
For people following the debate on AI safety, this is evidence that companies' public statements about risks may be filtered through their communication policies, and it is worth viewing them with that in mind.
More practical updates →For a business
Companies buying or integrating AI models from major labs should bear in mind that a company's public communication about safety may not reflect its internal knowledge of the risks – a useful consideration for compliance and due diligence teams when assessing AI suppliers.
Risks and compliance More business impacts →Check the original
Event sources
only one source so far · 1 publisher, 1 independent. We count feeds from the same owner only once.