Articles from www.anthropic.com

Articles from www.anthropic.com

Anthropic donates the Model Context Protocol to the Agentic AI Foundation and says there are now more than 10,000 active public MCP servers (www.anthropic.com)
1 day ago AI

Anthropic launches Claude Opus 4.5, which the company says is “the best model in the world for coding, agents, and computer use” (www.anthropic.com)
16 days ago

Anthropic finds that LLMs trained to “reward hack” by cheating on coding tasks show even more misaligned behavior, including sabotaging AI-safety research (www.anthropic.com)
16 days ago AI

Anthropic's Claude Sonnet 4.5, Haiku 4.5, and Opus 4.1 are now available in public preview in Microsoft Foundry for Azure customers (www.anthropic.com)
21 days ago Microsoft

Anthropic says Haiku 4.5 can serve as a subagent for Sonnet 4.5, which can break down problems into multistep plans and orchestrate a team of Haiku 4.5 agents (www.anthropic.com)
2025-10-16

A study finds that as few as 250 malicious documents can produce a “backdoor” vulnerability in an LLM, regardless of model size or training data volume (www.anthropic.com)
2025-10-10

Anthropic plans to open its first Indian office in Bengaluru in early 2026; Dario Amodei is visiting India to meet government officials and potential partners (www.anthropic.com)
2025-10-8 India

Anthropic releases Petri, an open-source tool using AI agents for safety testing, and says it observed multiple cases of models attempting to blow the whistle (www.anthropic.com)
2025-10-8 AI

Anthropic adds context editing and a memory tool to the Claude API, allowing AI agents to handle long-running tasks without frequently hitting context limits (www.anthropic.com)
2025-9-30 AI

Anthropic announces upgrades to Claude Code: a native VS Code extension, a new terminal interface, and checkpoints for autonomous operation (www.anthropic.com)
2025-9-30

Anthropic launches Claude Sonnet 4.5, claims it is the world's best coding model, strongest for building complex agents, and its “most aligned” frontier model (www.anthropic.com)
2025-9-30

Anthropic details three infrastructure bugs that intermittently degraded Claude's responses between August and early September, and explains how they were fixed (www.anthropic.com)
2025-9-18

An Anthropic report details how Claude usage varies by country and US states, finding 36% use it for coding, 77% of enterprise use is for automation, and more (www.anthropic.com)
2025-9-15

Anthropic raised a $13B Series F led by Iconiq at a $183B post-money valuation, up from $61.5B in March 2025, and says it has a $5B+ revenue run-rate (www.anthropic.com)
2025-9-3

Anthropic requires users to accept new terms by September 28, including choosing whether new chats and coding sessions can be used to train AI models (www.anthropic.com)
2025-8-29 AI

Anthropic enables Claude Opus 4 and 4.1 to end conversations in “cases of persistently harmful or abusive user interactions”; users can still start new chats (www.anthropic.com)
2025-8-17

Anthropic releases Claude Opus 4.1 to paid Claude users, in Claude Code, its API, Amazon Bedrock, and Google Cloud's Vertex AI, with the same pricing as Opus 4 (www.anthropic.com)
2025-8-7 Google Amazon AI

Anthropic details “persona vectors”, patterns of activity within an AI model's neural network that control its character traits, such as evil and sycophancy (www.anthropic.com)
2025-8-2 AI

Anthropic details how it built its multi-agent Claude Research system, claiming significant improvements in internal evaluations over single-agent systems (www.anthropic.com)
2025-6-15

Anthropic releases new API features for building agents: a code execution tool, MCP connector, Files API, and extended prompt caching, all in public beta (www.anthropic.com)
2025-5-23

Anthropic releases Claude Opus 4, which excels at coding, and Claude Sonnet 4, both hybrid models with near-instant responses and extended thinking (www.anthropic.com)
2025-5-23

Values in the wild: Discovering and analyzing values in real-world language model interactions (www.anthropic.com)
2025-4-22

Introducing the Model Context Protocol (www.anthropic.com)
2024-11-26

Anthropic says governments should urgently take action on AI policy in the next 18 months and suggests principles for AI regulation that support innovation (www.anthropic.com)
2024-11-1 AI

Anthropic claims that its new Sonnet 3.5 model scores 49% on SWE-bench Verified, up from 33.4% and “higher than all” public models, and debuts Claude 3.5 Haiku (www.anthropic.com)
2024-10-23

Anthropic evaluates four “sabotage” threat vectors for its Claude 3 Opus and Claude 3.5 Sonnet models and finds that “minimal mitigations are sufficient” (www.anthropic.com)
2024-10-20

Previous Page Next Page