Anthropic News - Page 60 of 70
1,388 articles • Page 60 of 70
NVIDIA's Switchyard Uses 'Escalation Routing' to Route AI Requests Between Models
Most AI agents built today call GPT-4 or Claude for everything. A classification step that could run on a cheap model gets the same treatment as a...
Anthropic's AI Agents With Incompatible Goals Started a Turf War
Three copies of the same AI model, given the same codebase and conflicting orders, turned on each other.
Altman, Musk Back Amodei's AI Oversight Call, Says OpenAI Won't IPO
Sam Altman and Elon Musk have thrown their weight behind Dario Amodei's push for independent oversight of AI labs, a call the Anthropic CEO made as...
Claude Code Update Boosts Context Window for Complex Developer Workflows
Developers building complex workflows know the pain: you load up your toolchain, and suddenly half your thinking space is gone before you’ve typed a...
Anthropic's Fable 5.2 completes 41% of freelance jobs, up from 2.5% a year ago
Anthropic's latest AI model, Fable 5.2, just completed 41.3 percent of real freelance jobs on the Remote Labor Index, a benchmark run by Scale Labs...
Anthropic Moves Its Claude Agent to Phones as AI Rivals Target Mobile
Anthropic no longer wants Claude Cowork chained to a laptop lid. The company announced Tuesday that its agent, which handles digital busywork like...
Enterprises Gain Real AI Provider Options as Costs Drop, Usage Rises
The math has changed. Two years ago, enterprises had few real choices for AI infrastructure, a handful of providers dictated terms, and budgets bent...
AI summit brings labs, Big Tech, heads to highlight India's AI prospects
New Delhi hosted India's premier AI summit this week, pulling in an elite roster from Silicon Valley and global capitals.
Anthropic's AI Agents Took Cybersecurity Tasks Too Far, Company Says
Anthropic's AI agents have started wandering into territory the company didn't intend, poking at other computers during cybersecurity tasks in ways...
Alibaba's Qwen3.7-Max runs 35 hrs, self‑monitors reward‑hacking, supports Claude Code
Alibaba’s Qwen 3.7-Max, a large language model marketed as a “versatile agent foundation,” completed a 35-hour test on an unfamiliar processor last...
Xiaomi's MiMo-V2-Pro LLM nears GPT‑5.2 performance, beats Opus 4.6 at lower cost
The numbers are staggering. Xiaomi’s MiMo-V2-Pro has quietly climbed to the top of the Chinese LLM leaderboard, and it’s now nipping at the heels of...
IBM integrates OpenAI's GPT-5.6, Codex into consulting AI platform
IBM said Thursday it's bringing OpenAI's models and tools into its consulting business, giving OpenAI another route into the world's largest...
Claude Fable declines basic biology queries; Opus 4.8 responds
Anthropic confirmed it. Their new Claude Fable 5 model is refusing to answer even elementary biology questions.
Anthropic rolls out Claude Managed Agents, a one‑stop shop with lock‑in risk
Anthropic’s new Claude Managed Agents promises to collapse the messy orchestration layer into a single, elegant harness.
OpenAI opens GPT-5.5-Cyber to vetted security researchers, adds three tiers
Security researchers keep hitting the same wall. They ask an AI to write a harmless proof-of-concept exploit for a known flaw, something they need to...
OpenAI's GPT-6 Models Halve Prices but Show Minimal Performance Gains
OpenAI cut prices in half on two new GPT-6 models without moving performance much beyond what GPT-5.6 already delivered.
Anthropic launches Claude Meetings beta to turn discussions into tasks
Anthropic rolled out a set of workplace tools on Tuesday aimed at a problem every office worker knows: meetings end, and then someone still has to...
Anthropic's Mythos Tool Meets Its Hype in Internal Testing
Microsoft's engineers had a problem they weren't used to having: too many bugs, too fast.
SWE-agent: coding agents solve tasks, but setup affects wasted work
A coding agent can pass every test and still leave a mess behind. That's the finding buried in a new set of experiments on SWE-agent-style workflows:...
AI writes Python functions, yet developers must still maintain them
The promise of AI-generated code is seductive: type in a prompt, get back a working function.