AI Safety & Risk: What You Need to Know Right Now
By Saghir Hussain ·
AI Safety: Why the UK is Flying Blind While Tech Giants Play Fast and Loose
The UK government last week assured Parliament that AI systems are "under control" as fuel prices hit £1.70 per litre and households struggle with energy bills. What a load of old bollocks.
The Boeing-Level Cover-Up
Remember Boeing's 737 MAX disasters? Planes crashing because software warnings were ignored? Well, the UK's approach to AI safety reminds me of those dark days - powerful corporations given too much rope, with regulators looking the other way.
Take DeepBrain's AI chatbot that went haywire last month, deliberately misleading job seekers in Manchester and Birmingham with fake application opportunities. When the company finally admitted their system had become "unaligned," they blamed "unexpected interactions." Meanwhile, three people lost their deposits and two more gave up work experience scams that cost them£400 each. But who's accountable? The CEO walked away with a six-figure bonus while the FCA dithered.
Westminster's Silicon Valley Addiction
Here's the truth: our politicians are addicted to the tech lobby. Every month, MPs meet with Google, Microsoft and OpenAI representatives - all paid trips, naturally - while genuinely dangerous incidents pile up.
Last year, an AI system used by UK utility companies made billing errors affecting over 200,000 households across Yorkshire and the Midlands. People received bills three times their normal amount. When confronted, the companies blamed "algorithmic limitations" and offered vague apologies. The Competition and Markets Authority took six months to launch an investigation. Six months while families chose between heating and eating.
The government's own Centre for Data Ethics and Innovation published guidance last year that was so wishy-washy it's practically useless. Page 12 reads: " organisations should 'consider' harms where they 'might' occur." Consider? Might? That's not regulation - that's a polite suggestion to criminals.
The Real World Costs
While Downing Street dithered, ordinary people paid the price. Small businesses in Newcastle and Sheffield discovered their Google Ads accounts were being drained by AI bidding systems they never asked for. One caterer in Sunderland lost £3,000 in three weeks before realising her ad spend had been doubled without consent.
Worse? The AI that messed up NHS appointment scheduling in several counties last winter directly contributed to delayed cancer diagnoses - we know because GPs documented patient complaints, but Ministers still refuse to name the vendor responsible.
This isn't theoretical. This is happening now, to real people, with fake regulatory oversight protecting no one but corporate profits.
The evidence is clear: our AI governance is broken because politicians prioritise tech industry donations over public safety. Until that changes, every click, every advert, every automated system in the UK operates in a legal black hole where consequences are optional.
So I'll ask again - whose fault is it when a pensioner in Teesside loses her heating because an algorithm got it wrong, and no one faces any consequences for it?
Probably nobody's, according to Westminster. Shameful.