UK Wants AI Rules — But Who Do They Actually Protect?
Here’s the tension I can’t shake: the same governments now demanding new AI laws to protect human rights are the ones who spent the last decade failing to enforce the laws they already had against surveillance capitalism. So when I read that a UK cross-party committee wants a new AI bill with an “independent oversight body” and “strong powers to hold AI providers to account,” I ask — accountable to whom, exactly?
The specifics are worth taking seriously. Public face-scanning, explicit deepfakes, and workers flagged for discipline by opaque algorithms are real harms happening to real individuals right now, not speculative superintelligence risk. That’s the useful part of this report. But the framing keeps sliding toward the abstract and existential — bans on “artificial superintelligence,” Anthropic calling for an industry-wide slowdown, Sam Altman and Dario Amodei publicly negotiating for antitrust waivers so the biggest labs can coordinate without competing. Read that again: the companies asking for restraint are also asking for permission to collude. That’s not safety regulation, that’s incumbency protection dressed up as caution.
Meanwhile, the AI Security Institute has no power to compel Anthropic or anyone else to submit models for testing — a voluntary system that only works if companies feel like cooperating. That’s the actual gap. Not a lack of grand legislative ambition, but a lack of teeth on the concrete, individual-level harms already documented.
Source: UK MPs and Lords call for new laws to tackle AI threat to human rights (The Guardian)
This matters now because the UK is drafting AI legislation while the loudest voices calling for restraint are the frontier labs themselves — a conflict of interest Cogtropolis readers should scrutinize. It’s a live test of whether “AI safety” regulation protects individuals from face-scanning and algorithmic discipline, or just cements the market position of a handful of companies.Source: Original article
