← Back to live feed1 story
Executives at OpenAI and Anthropic are privately simulating government responses to potential AI-driven crises, including scenarios where attempts to ban AI development fail. These simulations envision large-scale cyberattacks disabling banking, power, water, and internet access, with some scenarios assuming an event in 2027.
Sign in to suggest edits
Key sources
- SOURCE@jimvandehei“gaming out scenarios for a public and political revolt after a catastrophic AI event”x.com
- SUPPORT@kimmonismus“most likely a cyberattack disrupting banking, internet access, power or water”x.com
- SUPPORT@m_ccuri“assume Democrats in power will try -- and fail -- to shut down AI through bans, pauses or kill switches”x.com
- SUPPORT@andrewcurran_“wargaming scenarios for the aftermath of a catastrophic event in the next 6 - 12 months”x.com
- SUPPORT@niubi“lawyers will have a field day going after the executives personally for all the gains they made”x.com
- SUPPORT@jon_stokes“blame open-source AI, and get a kind of Patriot Act for Compute in place that locks down the open threat”x.com
- SUPPORT@teortaxestex“I guess they'll need bunkers after all”x.com
- SOURCEmarketbrief.now