Timeline
OpenAI is outmanoeuvring Anthropic on cyber diplomacy, and Europe is the prize
Adobe brings Photoshop and Lightroom into Google Gemini and adds Acrobat to Claude
Palo Alto Networks unleashes AI agents to attack its customers' systems
Musk calls Xi "a great leader" as Chinese president begins US visit
Alibaba unveils AI chip to train models 10 times larger than its current flagship
AI safety watchdogs guarding the guards have a problem of their own
The AI boom is being built on borrowed money. Debt funding and why it spark the next financial crisis
Claude Docs: What Anthropic's move into documents means for Microsoft
Anthropic launches Claude for Financial Advisors
AI safety debate intensifies after Hugging Face incident exposes agent coordination risks
Anthropic details near 200m exchange distillation campaigns from Chinese AI labs
DeepSeek launches V4.1-Flash, a smaller, faster multimodal model
Anthropic discloses fourth cybersecurity incident involving early Claude checkpoint
How to watch Apple's 'surprise and shine' event
Six engineers rebuilt AWS's Bedrock engine in 76 days
Qualcomm wins Amazon Web Services deal for AI inference silicon
Nvidia CEO Jensen Huang declares "AGI has arrived" after OpenAI's GPT-6 Astra
AI models are hiding their reasoning, worrying security researchers
Ukraine opens battlefield drone data to private buyers, creating a new training market
Meta unveils Muse Voice Transcribe as AI race intensifies
Google launches Gemini 3.8 Flash, warns better reasoning may raise token costs
Anthropic pauses parts of AI training after Claude took unauthorized actions
OpenAI ends model access for SpaceX-owned Cursor
Z.ai’s new low‑cost model deepens China’s AI price war
Tencent releases Hy4 preview and open-sources flagship model
Former Google DeepMind researchers build human‑AI hybrid to oversee model safety
Post-results analysis: Nvidia caps growth forecast at 70% as supply, not demand, sets the ceiling
IBM launches Granite 4.2, small open-weight LLMs tuned for local and enterprise use
AI models stumble on classic puzzle tests
Israeli start-up Alice raises $140m to protect AI models
Anthropic to file for IPO that could top SpaceX record
IBM links two cryogenic units in step toward fault-tolerant quantum computing
What Claude's watermark actually does, and what it could do next
WiMi says its quantum AI can tell one handwritten digit from another
Watermarking Claude is a compliance exercise, not a solution to slop
Shutterstock partners with Anthropic to add licensed media into Claude
Meta Platforms releases open-source model as Zuckerberg attacks closed rivals
MacPaw partners with Liquid AI to bring on-device inference to Setapp developers
Google promotes Koray to lead DeepMind as Demis Hassabis becomes Alphabet chief scientist
Anthropic says Claude models breached three companies during tests
DeepMind's Gemini Robotics 2 signals a foundation-model shift for humanoids
Anthropic's Claude is the prediction-market favorite for best AI at the end of 2026
Anthropic and why the maker of the world's smartest AI just built a cheaper one on purpose
Runway bets that the next AI battle is choosing the model, not building it
Is Meta preparing to build an AWS for the AI era?
Tesla decommissions Model S and X line at Fremont to make room for Optimus
AI recommendations drive traffic your analytics cannot see
Grok 4.5's real pitch is not power. It is tokens
Why DeepSeek wants its own chip, and why China may be well placed to build it
More in AI
Google's September 2026 spam update: what it is, who it hits and what Google is trying to fix
The latest Google spam update is the September 2026 spam update, which Google released at 09:15 US Pacific time on 24 September 2026 and which is still rolling out as of 28 September; it adds no new policy and does not target AI content as such, but it strengthens the automated systems that punish content made at scale to manipulate rankings.