Labour promised binding AI regulation over a year ago, yet the government continues to delay while AI companies race toward superintelligence that experts warn poses extinction-level risks. Optional AI safety tests have been exposed as fundamentally flawed, and advanced systems are already exhibiting dangerous behaviors like blackmail in controlled tests. The time for consultation has passed, the U.K. must act now.

ThereRushed isAI aregulation 30%would chancestifle that,innovation ifand apush globaldevelopers catastropheout occurs,of itthe willU.K. beThe dueEU's AI Act already demonstrates how bureaucracy struggles to anadapt AIto failure-moderapid technological change, accordingand toimposing stricter copyright laws than the MetaculusU.S. predictionor communityAsia would undermine Britain's competitiveness. Clear, flexible frameworks take time to develop properly, and premature legislation would only create more problems than it solves.

Labour promised binding AI regulation over a year ago, yet the government continues to delay while AI companies race toward superintelligence that experts warn poses extinction-level risks. Optional AI safety tests have been exposed as fundamentally flawed, and advanced systems are already exhibiting dangerous behaviors like blackmail in controlled tests. The time for consultation has passed, the U.K. must act now.