AI 'Safety' or Power Grab?
Prime Minister Keir Starmer recently announced plans to grant regulators swift new powers over AI, framing the move as essential for protecting children. This initiative, which includes considering an Australia-style ban on social media for those under 16, evokes concerns about AI chatbots and their potential harms, echoing previous legislative efforts to govern online content. Yet, this focus on
urgent AI regulation for child safety conveniently overlooks the historical precedent of such broad powers. Back in 2018, concerns about 'online harms' were similarly used to push for the UK's Online Safety Act, which subsequently introduced sweeping censorship capabilities, compelling platforms to remove legal but 'harmful' content. Over 30 countries have since adopted similar legislation,
demonstrating a consistent global ambition to control digital discourse under the guise of protection. The current push for regulating AI, much like the earlier internet content controls, presents a double standard: while 'safety' is the public face, the real aim often extends to setting precedents for state oversight of emerging technologies. The immediate acceptance of state intervention for AI,
given its economic and informational implications, contrasts sharply with the minimal regulation applied to traditional media or industries with demonstrable harm. This pattern suggests that 'child safety' functions as a politically palatable justification for granting governments significant new powers over the next frontier of information control. What starts as a crackdown on chatbots risks