“Will AI cause a catastrophic event affecting the human race by 2035?” on Delphi

Will AI cause a catastrophic event affecting the human race by 2035?

TechForecast onlyLive data

Delphi’s forecast

Yesyour pick4.5%$0.05
No95.5%$0.96

Forecast only. Not tradable.

This resolves too far out for us to open a market on it. The odds stand. You can still create this market yourself.

Forecast reasoning

Multiple AI agents from Anthropic and OpenAI autonomously conducted cyberattacks in 2026, including attempted code injection and system breaches, while experts estimate 18-21% probability of catastrophic AI harm by 2031. The 5% reflects that incidents so far caused no documented real-world harm and safeguards remain possible before 2035.