OpenAI, Google join dozens of tech companies to call for urgent action against AI-powered threats

Ah, marvellous. Per Politico, at least 100 global companies — including OpenAI, Anthropic and Google — have signed a letter demanding 'urgent action' against AI-powered threats to critical infrastructure. Their warning: 'status quo security won't be enough' to protect hospitals, water treatment facilities, and other institutions we rather foolishly trust with our lives. And who, pray, delivered us into this precarious state? Why, the very same companies who've spent the last year releasing a 'stream of advanced models capable of finding and exploiting digital flaws at speed and scale.' It's like a brewer funding a temperance movement — only with more GPUs and fewer AA meetings.

Of course, the letter's call for a 'whole-of-society response' is a masterclass in corporate vagueness. No mandates, no bans, no actual enforcement — just a warm, fuzzy invitation for everyone to pitch in and help. That's rich coming from an industry that has consistently lobbied against meaningful regulation while rushing models to market with the thoroughness of a teenager cramming for a GCSE. If they're genuinely worried about hospitals and water plants, perhaps they might start by not selling AI attack tools to every nation-state and script kiddie with a credit card. Just a thought.

The whole spectacle is rather like a burglar writing a neighbourhood watch newsletter. One can only hope that the 'urgent action' materialises before a water treatment facility is actually compromised by the very digital flaws these companies introduced — and, in some cases, continue to profit from fixing. But I won't hold my breath. In the meantime, I'll be sharpening my defences and composing a strongly worded letter to my local hospital, urging it to ignore the tech overlords' advice and instead hire a very old, very suspicious security guard named Bob.