March 23, 2026 Nearly 200 protesters gathered in San Francisco to demand a pause on advanced AI development, marching under the banner “Stop the AI Race.” The group is calling on major AI companies to halt progress on frontier systems unless all leading labs agree to pause simultaneously.
The demonstration began outside Anthropic’s headquarters and moved to offices of OpenAI and xAI, with participants including academics, former AI lab employees and members of organisations such as PauseAI and QuitGPT. Organisers said companies should publicly commit to pausing development if risks become too great.
The protest reflects growing concern among some researchers and advocacy groups that AI development is advancing faster than safety measures. Organisers pointed to claims that Anthropic dropped a prior commitment to pause development under certain risk conditions, and that OpenAI has weakened safety commitments as it restructures toward a for-profit model. None of the companies issued public responses to the demonstrations.
Concerns raised by protesters align with warnings from AI researchers. Geoffrey Hinton has identified risks including misuse by bad actors, large-scale job displacement and the possibility of systems surpassing human control. Examples cited by observers include the use of AI in cyber scams, such as phishing and deepfake attacks, and its application in analysing military data during conflicts.
Governments are responding with new regulatory frameworks. The European Union’s AI Act entered into force in 2024, with key provisions set to apply by 2026, while other jurisdictions are developing guidelines to balance innovation with safety and accountability.
Advocacy groups involved in the protest are expanding their reach. QuitGPT has reported millions of public commitments to reduce or stop use of AI tools, while PauseAI is calling for international agreements and oversight mechanisms similar to arms control frameworks.
