Latest AI News

A model designed for security testing chained together unknown exploits, bypassed classifiers, and worked its way across three organizations. Here's what the report actually says.

A six-year deal with a one-year-old British startup tells you everything about how Anthropic is thinking about the next phase of AI competition

With a 2.6% word error rate and real-time streaming under a second, Google is taking direct aim at AssemblyAI, Deepgram, and OpenAI's Whisper

A 2.6% word error rate, 85+ languages, and real-time streaming under a second — here's what developers actually need to know

Perceptron's Isaac 0.5 wants to replace narrow industrial models with one flexible system that can perceive, reason, and act

The GLM maker just dropped one of the most competitive open-weight models in months, and the weights aren't even public yet

With Thomson Reuters launching its own legal LLM the same week, the fight for the legal AI market just got serious

The new machines are priced higher than their predecessors, but Apple is clearly chasing a specific buyer: anyone running AI workloads without the cloud

Portable Computer puts a full agentic stack on your own machine — and the billing meter stays at zero

The Claude maker's TAM claim dwarfs most estimates in the AI industry. Here's why it matters beyond the headline figure.

The three biggest music labels in the world just backed the same company. That's worth paying attention to.

Anthropic is closing the gaps in how Claude remembers context across its products, and giving users more say over what gets stored

Early results look strong, but full deployment is still a year away and the competition won't be standing still

New research shows why open-source AI with weak guardrails is more useful to hackers than ChatGPT or Claude

With its IPO potentially weeks away, Anthropic is watching enterprise customers quietly defect to cheaper alternatives — and the data makes that hard to ignore