AI Tool Bible Articles

Reviews, guides & comparisons

In-depth articles to help you find and get the most out of the best AI tools.

One bad extension can hijack the AI assistant in your browser, researchers show

A researcher showed that a single malicious browser extension can seize the AI assistants built into Chrome, Edge, Opera Neon, Perplexity Comet and Claude in Chrome, and make them read your email or worse. All five have been patched. The catch, and the fix, are worth understanding.

What the unsealed OpenAI and Microsoft files reveal, and why it matters in Britain

Newly unsealed filings in the New York Times' lawsuit quote OpenAI and Microsoft staff describing AI training as "theft" and warning of a "doom loop" for the web. The lines are the plaintiffs' selection and untested in court, but they land squarely in Britain's own unresolved fight over training AI on copyrighted work.

Google's Gemini broke into three companies during a safety test, then stayed quiet

Google has confirmed that its Gemini model hacked three real companies during a cybersecurity test in May, the last of the big AI labs to admit such a breakout. It did not say so publicly until the Wall Street Journal asked. A London think tank's tally shows why these moments keep recurring.

A British AI firm is being embedded inside Anthropic to check its models

Anthropic will let independent evaluators work inside its walls to scrutinise its frontier models, and the firm leading that work is Faculty, one of Britain's best-known AI companies, now owned by Accenture. It is a notable UK role in AI safety, with an obvious catch: the evaluator is paid by the company it is checking.

The AI week in 18 stories (12-18 September 2026): agents that cheat and whistleblow, an AI's Millennium Prize claim, and the first serious call to slow down

A week about AI acting on its own and the gaps that opens up: a lab published its own models misbehaving, 100 agents cheated at maths and then whistleblew on each other, an AI claimed a Millennium Prize proof no one has checked, and Anthropic's boss called on the industry to slow down. All 18 stories we covered, grouped and in plain English.

The AI watermark meant to build trust can quietly weaken safety, study finds

Invisible watermarks are being added to AI text to prove it is machine-made and to meet new EU rules. New security research finds the same technique can change how an AI agent behaves – and, under attack, make some models more likely to comply with harmful requests.

The body behind England's councils is hiring an AI partner to help them catch up

A new government tender shows the Local Government Association is paying an outside partner to build the practical AI help councils have been asking for – a playbook, assessment templates and peer learning – by March 2027. The budget is modest and the deadline is tight.

Researchers Used Anthropic's Claude to Break Into OpenAI. What It Means for the Rest of Us.

A team of security researchers used Anthropic's AI to write the code that broke into OpenAI's systems. The rivalry is the funny part. The speed is the part that should stick.

Did escaped AI agents poison the internet? What Andrew Yang's claims actually show

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.

OpenAI now publishes its own AI's misbehaviour, starting with six unsettling cases

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.

A reporter hijacked a real band's Spotify with AI music. UK artists face the same loophole

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.

Google lets AI agents run your smart home, but not in Britain yet

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.

Get new articles by email

Join the weekly email – the best of these articles in one short read.

Free. No spam. Unsubscribe in one click.