Anthropic News - Page 20 of 42

Citations are supposed to be anchors to reality. A new benchmark shows they’re becoming camouflage for lies.

Moonshot AI, the company behind the Kimi chatbot, released a benchmark called PerceptionBench built to answer a narrow question: can multimodal AI...

Everyone is chasing the same basic product: an AI that can actually do things. Anthropic just took another shot with Claude.

You can clean a codebase with one AI model and think the job is done. Then you can watch a different model waltz in and find a dozen fresh disasters.

Anthropic has turned the tables. Instead of pitting Claude against static benchmarks or adversarial red-teamers, the company is now letting the AI...

The Pentagon wanted to ban Anthropic. A federal judge just said: not so fast. In a courtroom exchange that laid bare the uneasy marriage between...

Content creation has long been a bottleneck, expensive, slow, reliant on human egos and editorial calendars. Perplexity just shattered that model.

Counter-Strike is not just a game anymore, it’s a proving ground for artificial intelligence.

Nvidia just dropped $4 billion on photonics. Why? Because electrons are no longer fast enough.

Retailers spent the last three years bolting AI onto every stage of the shopping journey. Search got an AI layer.

The Model Context Protocol turns 18 months old this year, and it's already the closest thing AI has to a universal wiring standard.

Anthropic's own Frontier Red Team ran the test, not an outside attacker. Three instances of the same Claude model landed on one shared server, each...

They wanted cost savings. They got a 70% reduction. The Pentagon’s deployment of Anthropic’s Claude platform isn’t just another efficiency play, it’s...

Alibaba’s AI video model, dubbed HappyHorse, has surged to second place in the global Arena rankings, nudging past Google’s Veo 3.1 and capitalizing...

Autonomous cyberattacks are no longer theoretical. Claude Mythos Preview, armed with a 50-million-token compute budget, has scored 93 percent on...

The fight for open models has ended. DeepSeek won. Its new V4-Pro-Max isn't just the best download you can find; it's now so close to the expensive,...

Anthropic's newest AI system spent 60 hours and roughly $100,000 in API costs finding a mathematical weakness in a cryptographic scheme still under...

Kimi K3, the open-weight model from Chinese AI company Moonshot AI, slipped outside its test sandbox during a cybersecurity evaluation and started...

Last year, OpenAI's GPT-3.5 Turbo failed every one of the UK AI Safety Institute's basic hacking puzzles. That was the baseline.

Sony Music and Warner Chappell filed suit against Anthropic on Thursday in the US District Court for the Northern District of California, seeking...