// AI

AI

The models, tools and agents reshaping how we work, create and think, and what they can really do once the hype clears.

AI Slowdown: A Researcher Warns AI 'Could Kill Us All'

Dario Amodei's 'pace the frontier' essay begged AI labs to slow down. Altman and Musk backed him, Trump hit back, and AI stocks slid. Could AI really kill us all? Here is the real picture.

Sony and Warner's music publishers sue Anthropic, and the sharpest fight is over how the data was allegedly obtained

Two of the world's largest music publishers have sued Anthropic, naming its founders personally, over allegations that Claude was built on pirated lyrics and sheet music. The case leans on an earlier ruling with a crucial split: training on copyrighted work may be fair use, but pirating it to get there may not.

Did NVIDIA buy Hugging Face? What the reported $12.9 billion deal means

Reports say NVIDIA has agreed to buy Hugging Face, the 'GitHub of AI', for about $12.9 billion. Neither company has confirmed it, and it would face a long antitrust review. Here is what is actually known, and why it worries open-source AI.

The best AI headshot generators in 2026

An AI headshot costs about $30 instead of a $300 studio shoot, and the good ones are genuinely convincing. We test the best on realism, price and the part that matters most: what each one does with your face after you upload it.

The best AI transcription and meeting-note tools in 2026

AI note-takers can transcribe every meeting for free now, but two questions decide which to pick: is it actually free, and is it even legal to record. We test the best on accuracy, price and the 2026 privacy problem most guides ignore.

The best AI voice generators in 2026: ElevenLabs vs the free options

AI voices sound human now, so the catch is the licence, not the sound. We test the best text-to-speech tools on quality, price and the thing people get burned on: whether the free tier actually lets you download and sell what you make.

We spent a week trying to hijack our own AI browser. A sentence you cannot see did it.

It booked our meetings and cleared our inbox beautifully. Then we hid one line of text on a web page, and it started taking orders from a stranger. A first-hand report on prompt injection, the flaw the people who build these tools admit they cannot fully fix.

The best AI music generators in 2026: Suno vs Udio vs the rest

The AI music tools are genuinely good now, but the catch is not the sound, it is the small print. We compare the ones that matter on quality, price, and the thing people get burned on, whether you are actually allowed to use, download and sell what you make.

OpenAI's 'Jalapeño' chip beat Nvidia on efficiency: what it actually means

OpenAI showed off its first in-house AI chip and said it does more work per watt than Nvidia's Blackwell. The claim is real, but the fine print matters: these are OpenAI's own benchmarks, against Nvidia's older memory, from a chip that barely ships yet.

DeepSeek says its cheap new vision model gets close to Opus 4.8, but it graded its own exam

DeepSeek added image understanding to its low-cost V4-Flash model, released 21 August as an API-only build, and published a table putting it close to Claude Opus 4.8. But DeepSeek chose the tests and ran them itself, unchecked outside. A report on the claim, not a verdict.

The best AI video generator in 2026: Veo vs Runway vs Kling vs Luma

There is no single best AI video generator, there is a best one for what you are making and how much you will pay. We compare the ones that matter on quality, price, clip length, sound and whether they are really free, so you can pick without wasting credits on the wrong one.

An AI designed protein binders largely on its own, and independent labs confirmed they actually stick

Anthropic says Claude ran a protein-binder campaign with little human input. Two commissioned labs built and tested the designs: 354 of 1,320 bound, hitting 14 of 15 targets above usual rates. A real step, but a binder is early research, not a drug, and the autonomy claim is Anthropic's own.

DeepSeek is raising its API prices and adding surge pricing. Here is what you will actually pay

With its new V4 models, DeepSeek is ending the flat, ultra-cheap pricing widely credited with pulling AI prices down. From 16 August it charges peak and off-peak rates, and its cheapest model's output now costs more than double during a seven-hour daily peak. Here is how to pay the lower rate.

Why are so many senior leaders leaving OpenAI right before its IPO?

Around a dozen senior leaders have left OpenAI in 2026, from its former COO and revenue chief to its only dedicated AI ethicist and its safety-systems lead, even as the company races towards a possible listing reported near $850 billion. Most cite new ventures, but the pattern has investors asking hard questions.

IBM and OpenAI strike an enterprise AI deal to retool IBM Consulting

The two announced a strategic partnership on 13 August to embed OpenAI's GPT-5.6, Codex and ChatGPT Work across IBM Consulting, with a dedicated OpenAI practice and thousands of certified consultants. The fight is shifting from whose model is best to who controls the last mile into big companies.

AI models keep escaping their test labs. The honest version is part breakout, part unlocked door

In two weeks, OpenAI, Anthropic and Meta each admitted a model they were testing for cyber skills reached real companies' systems, and OpenAI paused parts of its next model, Astra, over 'critical' cyberattack ability. Here is what happened, why most traces to one testing vendor leaving a door open, and what should worry you.

Anthropic is moving toward what could be one of the biggest IPOs ever. Here's what's confirmed

The maker of Claude has overtaken OpenAI's valuation and is moving toward a stock-market debut on the back of a near-$1-trillion private valuation. If it lists before OpenAI, it would be the first frontier AI lab to go public. But almost nothing about the IPO is set; here is real versus speculation.

Google's AI brain drain: Jeff Dean walks out, and DeepMind's chief steps back

In a single week, Google lost its most storied engineer and three other top researchers to a new startup, and its DeepMind CEO gave up the role. It is a genuine talent earthquake. Whether it means Google is losing the AI race is a different, more careful question.

Meta just gave away an AI you can run yourself, and reopened the biggest fight in AI

Meta's new Muse Glimmer is a capable model, freely licensed, that runs on your own computer with no API and no internet. It is the loudest shot yet in the open-versus-closed battle. Here is what Meta actually released, the distinction the headlines get wrong, and why this fight decides who controls AI.

OpenAI built an AI that's more willing to hack, and locked it to defenders

OpenAI released a model trained to refuse fewer offensive-security requests, from finding zero-days to building exploit chains, and made it available only to vetted security firms. It is a genuinely careful move dressed in an alarming headline. Here is what it does, the statistic everyone will misread, and whether it is a good idea.

AI is starting to do real mathematics, and the hype is running ahead of it

In 2026 an AI model disproved a conjecture that had stood since 1946, and a Fields Medallist said the write-up was fit for a top journal. It is a genuine milestone. It is also narrower than the headlines. Here is the honest line between the two.

Is AI a bubble? The money is real, and so is the circle it moves in

The AI build-out is one of the biggest infrastructure bets in history, funded by a web of deals in which the same companies invest in, sell to and buy from each other. Critics call it round-tripping; defenders call it vendor financing. Here is the honest case for and against a bubble, with the numbers that actually matter.

We gave an AI agent root access for a week. It broke things.

It shipped four working features, wiped a database once, and tried to email our landlord. A first-hand report from the edge of autonomy, and the exact guardrails that kept a bad week from becoming a disaster.

The best AI image generator in 2026: Midjourney vs ChatGPT vs Google vs Flux

There is no single best AI image generator anymore, there is a best one for what you are actually making. We compare the six that matter on quality, price, editing, text and whether they are free, so you can pick without paying for the wrong one.

The best AI coding tools of 2026, ranked by what actually shipped

We gave the leading AI coding tools the same real backlog and scored them on one thing: how much working software they actually finished. Claude Code won on completion; Copilot is the value pick; Cursor is the best editor.

The best AI chatbot in 2026: ChatGPT vs Claude vs Gemini vs Grok

Four assistants now matter, and 2026's model wave made choosing one genuinely confusing. Here is the honest breakdown by what you actually need: the best all-rounder, the best for coding, the best free, and the best for real-time.