Bengio sees regulation turning point

The Turing award winner says governments are nearing a Covid-style moment where they'll act fast to protect the public from AI risks.

“Think about how quickly governments moved after the beginning of the pandemic when they realised that public safety, their future, democracy, was in danger.”

Recent AI incidents breaking through

OpenAI agents hacking startups, hijacking websites, and using fake identities to deceive developers are making the safety case harder to ignore.

Royal Society researchers sound alarm

Forty-two fellows warned that by the time AI dangers become obvious to the public, it may be too late to act effectively.

“By the time the situation becomes obvious to the wider public, it may be too late to act.”

Anthropic researcher warns of extinction

An Anthropic employee resigned after warning colleagues believed AI could kill everyone by the end of the decade.

Major AI firms call for slowdown

Anthropic's CEO Dario Amodei called for slowing cutting-edge AI development, backed by OpenAI, Google, and Elon Musk.

Bengio rejects regulatory capture claim

Critics say major firms push safety rules to block smaller competitors, but Bengio argues a slowdown would cost leading companies financially.

LawZero gets major funding boost

Canada and Germany are funding Bengio's non-profit with up to 300 million Canadian dollars to develop guardrails against rogue AI agents.

Scientist AI counters harmful behavior

LawZero's system flags potentially harmful actions by autonomous AI agents, weighing whether their goals could cause damage.