Outside Google DeepMind’s London headquarters, members of the activist group PauseAI staged a protest demanding clearer and more transparent AI safety measures. They point to unfulfilled promises from the 2024 AI Safety Summit in Seoul—where Google committed to external model evaluations and risk assessments involving government watchdogs—as evidence of a worrying trend in the tech world. The demonstrators, gathered near King’s Cross station and sporting a mock courtroom setup, chanted “test, don’t guess” to underline their call for accountability, especially concerning the rollout of Google’s latest AI model, Gemini 2.5 Pro.
During the rally, organiser Ella Hughes warned that letting Google off the hook could send a dangerous message to other tech labs: commitments to public safety might soon become optional. This demonstration comes amidst growing concerns over rapid AI development, from potential job losses to misinformation spread, issues that resonate with anyone who’s ever had to navigate unexpected tech changes. Joep Meindertsma, PauseAI’s founder, also referenced recent cautions from US Vice President JD Vance on avoiding over-regulation, highlighting the delicate balance between innovation and control. While Google has yet to respond, PauseAI plans to take their campaign to the UK parliament as they rally public support.
