Anthropic CEO Dario Amodei On Why He Left OpenAI, Viral AI Summit Clip And What's Next For AI

There was a moment at India's AI summit where Anthropic CEO Dario Amodei and OpenAI's Sam Altman appeared to dodge holding hands on stage and the internet noticed instantly. Speaking in a Bloomberg interview, Amodei was asked point-blank about that viral clip, his OpenAI exit and what came out was a mix of logistics, mild exasperation and a much bigger conversation about trust in the AI industry.What Actually Happened On That Stage?According to Amodei, the whole thing boiled down to chaos backstage rather than any deliberate snub. He said the summit was disorganised, with the lineup of speakers getting rearranged at the last minute before organisers including Indian Prime Minister Narendra Modi, abruptly told everyone on stage to hold hands for the photo.

Amodei pushed back on the idea that this was an India-specific problem, noting that pretty much every major international summit involving heads of state tends to run this way. When pressed on why everyone else managed to link hands while he and Altman seemingly didn't, Amodei didn't have a tidy answer. He simply said, "I don't know what to tell you," before circling back to how sudden and chaotic the instruction felt in the moment."Some Players Are More Trustworthy Than Others"During theinterview, Amodei acknowledged that cooperation across the AI industry is essential, particularly as companies race to develop increasingly capable models. However, he also suggested that not every player in the sector inspires the same level of confidence.According to Amodei, the industry's more reliable actors need to work together to establish standards that others are effectively compelled to follow. He argued that some organisations do not voluntarily adopt responsible practices unless broader industry pressure leaves them with little alternative.While he did not directly name specific companies during that part of the discussion, the comments are likely to fuel ongoing debates about competition, governance and safety within the AI sector.Why He Really Left OpenAIThen came the question that's been Silicon Valley gossip for years: what actually went down when Amodei left OpenAI to start Anthropic? He didn't dodge it, but he also wasn't interested in turning it into a soap opera. Amodei said there were genuine, valid disagreements over safety, but stressed that disagreements alone weren't the real reason he left, after all, he noted, people at Anthropic disagree with him and with each other too, and that doesn't break things apart. What actually mattered, he said, was a deeper erosion of trust, moments where he felt people's stated values didn't match their actions, where honesty seemed to be missing, and where the motivations being claimed didn't line up with what he was actually seeing. He described noticing what he called disturbing patterns of behaviour and dishonesty, the kind that make it genuinely difficult to keep working with or trusting a company.Amodei said that when two people or organisations don't share the same vision and don't trust each other, there's not much point arguing about it, you simply go your separate ways and let each side build what it believes in. He said he's completely at peace with that outcome, framing it as a question that will eventually get answered by the market and by public opinion rather than by debate.