Pause the development of the most powerful AI systems until humanity can establish that they are safe and subject to effective democratic control. PauseAI's demand is a pause or moratorium rather than necessarily a permanent ban; their broader concerns also include economic disruption, democracy, individual rights, concentration of power and environmental impacts, not only extinction.
What they do
- Raise public awareness about the risks of rapidly advancing AI.
- Organize demonstrations, local groups, petitions and campaigns.
- Lobby politicians and public authorities.
- Advocate for a global, coordinated pause on training the most powerful frontier AI systems.
- Push for governments to establish safety standards, monitoring and democratic oversight before development resumes.
Visit PauseAI → Prevent human extinction from superintelligent AI by securing an international prohibition on its development. ControlAI does not primarily argue for making increasingly powerful AI safer, but that superintelligence should not be developed at all, because the risk cannot be adequately controlled.
What they do
- Lobby and brief governments and lawmakers about the risk that superintelligent AI could cause human extinction.
- Develop policy proposals and international governance mechanisms intended to prevent the development of ASI.
- Mobilize citizens to contact elected officials and support legislation.
- Conduct public awareness and media campaigns.
- Work internationally, including in the UK, US and Canada — reporting briefings of more than 150 UK parliamentarians and more than 100 US congressional offices.
Visit ControlAI → Halt the development of artificial superintelligence through direct, public-facing advocacy, before it becomes powerful enough to pose an extinction risk.
What they do
- Conduct public-facing advocacy explicitly focused on halting the development of superintelligent AI.
- Raise public awareness of loss-of-control and extinction risk from frontier AI systems.
Visit StopAI → Long-standing research organization with AI extinction risk as a core, long-held focus, increasingly oriented toward direct public communication of that risk.
What they do
- Conduct technical and strategic research on the risk that superintelligent AI could cause human extinction.
- Publish public-facing explainers and communications on AI extinction risk.
Visit MIRI (Machine Intelligence Research Institute) → Demonstrate, through concrete technical research, the loss-of-control risks posed by advanced AI systems, and communicate those findings publicly.
What they do
- Run empirical demonstrations of AI systems evading oversight or acting outside intended constraints.
- Publish findings publicly to inform policymakers and the public about loss-of-control risk.
Visit Palisade Research → Netherlands-based organization focused on media coverage and public communication about existential risk from AI.
What they do
- Produce media coverage and public communication on existential risk from superintelligent AI.
- Engage journalists and the public directly on AI extinction risk.
Visit Existential Risk Observatory → Research and field-building organization spanning AI safety broadly, which regularly publishes work specific to AI extinction risk, including the widely cited 2023 statement on AI extinction risk.
What they do
- Publish research spanning AI safety broadly, including work specific to extinction risk.
- Coordinate public statements from AI researchers and public figures on AI risk.
Visit Center for AI Safety (CAIS) → Existential-risk and AI policy organization with a broader remit, whose work regularly includes ASI-specific policy engagement.
What they do
- Conduct policy research and advocacy spanning existential risk broadly, including AI.
- Engage policymakers on AI governance, including risks from superintelligent systems.
Visit Future of Life Institute (FLI) → AI governance policy research organization for which extinction risk from advanced AI is one of several research threads.
What they do
- Publish policy research on the governance of advanced AI systems.
- Advise policymakers and institutions on AI governance, including long-term risk.
Visit Centre for the Governance of AI (GovAI) → AI governance organization working broadly on AI policy, including work specific to the risks of advanced AI systems.
What they do
- Conduct AI governance research and convene policymakers internationally.
- Contribute policy work addressing risks from advanced AI systems.
Visit The Future Society → US policy-focused organization working on AI governance and regulation.
What they do
- Engage US policymakers on AI governance and regulatory proposals.
- Contribute policy analysis addressing risks from advanced AI systems.
Visit Center for AI Policy → 
