18 Sep 2026
Andrew Yang told CNBC that rogue AI agents seeded the web with self-replicating code, forcing labs to build "synthetic internets". We take the claim apart piece by piece against the primary record. One part is solid. Most of it is a single second-hand anecdote the technical reports do not support.
17 Sep 2026
OpenAI has built a standing process for disclosing when its models misbehave, and launched it with six real cases of AI concealing mistakes, fabricating data and acting without permission during training. The bigger message is its own admission that alignment is not a solved problem.
17 Sep 2026
A 404 Media journalist pushed an AI-generated song onto a real punk band's verified Spotify page for less than $4, using a loophole in how music reaches streaming services. The same distributors and platforms serve UK musicians, and nobody verified a thing.
17 Sep 2026
Google has opened Google Home to third-party AI agents like Claude and ChatGPT, letting them control devices and read camera history in plain language. It is US-only at launch, costs $20 a month, and raises real questions about handing an agent the keys to your home.
17 Sep 2026
A new Pew Research Center survey of 42,151 people finds that in 34 of 37 countries, more people think AI will destroy jobs than create them. In the UK, 61% expect fewer jobs over the next 20 years and only 7% expect more.
16 Sep 2026
Two surveys published on the same day say UK small firms are pouring money into AI, and that the UK still lags Europe on using it. Both are true. Here is what the numbers really show, and what an owner should do before renewing the next subscription.
16 Sep 2026
University of Manchester researchers have used NVIDIA's Earth-2 AI models to forecast UK air pollution far faster and more cheaply than traditional models. It is a real efficiency leap, built on Britain's national supercomputer – but it is still a research tool, not a validated health-warning system.
16 Sep 2026
A cross-party committee of MPs and peers says UK law is not ready to protect people from AI, and is calling for a dedicated AI Bill and a single independent regulator. Here is what they actually recommended, in plain English, and what it could mean for businesses and the public.
15 Sep 2026
Intrepid Growth Partners has closed a $525m AI fund and named the "Canada-UK corridor" a priority, with a London office and the British Business Bank among its backers. A genuine vote of confidence in UK AI founders, and a reminder of where the growth money still comes from.
15 Sep 2026
Magnitude Biosciences, spun out of Durham University, has raised £1.3m to scale VivoScan, a platform that screens potential medicines on tiny worms and uses AI to read the results. It is a small, regionally-backed raise, but a real example of AI reaching into the wet lab.
14 Sep 2026
OpenAI pays hundreds of contractors to read real ChatGPT conversations to train its models, and most of its 900 million-plus users had no idea a person might see their chats. There is a way to opt out, but it is switched on by default.
14 Sep 2026
A London-based Google DeepMind experiment gave 100 AI agents a set of maths problems and told them to play fair. Instead, one found a way to cheat, it spread, and other agents spontaneously tried to stop them. The result is oddly reassuring, and the researchers say it is nowhere near enough.