
Anthropic AP Licensing Deal: First Publisher Pact, Unconfirmed
Adweek says Anthropic signed its first publisher license, with the AP, before the midterms. Neither side has commented. Here is what is confirmed.
New AI, explained the day it ships. Frontier model releases, agent guides, Claude Code deep dives, MCP servers, AI careers, and global policy — 4,100+ articles from the instructors who've taught 350,000+ students.
Browse by topic →
Adweek says Anthropic signed its first publisher license, with the AP, before the midterms. Neither side has commented. Here is what is confirmed.

Baseten Grid claims big coding savings vs Opus 5.5. We checked the 72% figure, the 0.9-point quality gap and what the preview really includes.

Microsoft published a Copilot Managed Runtime repo: Vite 8 template, ms CLI and a coding-agent plugin. Setup, 0.1-credit billing and limits explained.

Russia floats underground data centers after Ukrainian drone strikes. Ministry guidance, Rostelecom-Norilsk talks and mine proposals: confirmed vs unconfirmed.

El-Sayed AI plan: pause AI development, FDA-style agency, kill switch, public board seats. What the plan says, what is unverified, who else proposes it.

Agent Autoloop runs Claude Code and Codex from cron, GitHub issues and HTTP calls, and sends approval cards. Install steps, rules, limits and security risks.

Sam Altman made an unannounced Korea trip. SK Hynix rose 2.4% on reports of HBM talks it could not confirm. Here is what is known and what is not.

Use decision models in Amazon Bedrock AgentCore Evaluations: noul, choice and score questions, Jev vs Strands Decider 2B, Lambda setup and validation.

Anthropic used Claude to cut vault purification from 3 days to 90 minutes and get a first cryo-EM model in 34 hours. Results, errors, open questions.

Gurman reports Apple lowered some engineers’ daily AI cap from $300 to $200, with $100 next. Claimed vs verified, and what it means for AI costs.

Arm says it holds about 50% of CPU compute at top hyperscalers, and now "over half" in headlines. What the SEC letter says and what no one has audited.

Atmanity says Atmee v2 runs live avatars on one GPU, 50x faster and 10-100x cheaper than Wan 3.0 and Seedance 2.5. See what is verified.

Axiom released a 16M-line Lean formalization of the Stacks Project, split into human-reviewed and AI-written tiers. What it is and what to check.

Axion 1.0 trains an agent model with 18K examples, two SFT stages and RL. Scores, costs, license and what the open recipe really includes.

Claude Code product head Cat Wu left Anthropic before its IPO. Here is what Business Insider reports, what Anthropic confirmed and what to watch.

ChatGPT Images 2.6 reportedly makes animated GIFs with speed and frame editing. See what is confirmed, what is not, and how to test it today.

Claude Opus 4.6 and Sonnet 4.6 watermark only at default sampling settings. Four more older models follow Nov 16. What is confirmed and what to do.

Cloudflare now shows which personal AI agents visit your site and lets you allow, block or ad-gate each one. What changed and the catches.
Page 1 of 229 · 4120 articles