Anthropic AI Updates: May 29, 2026

1. Anthropic Raises $65B Series H at a $965B Post-Money Valuation

Anthropic. Anthropic closed a $65 billion Series H at a $965 billion post-money valuation, led by Altimeter Capital, Dragoneer, Greenoaks, and Sequoia Capital, with co-leads including Capital Group, Coatue, D1 Capital Partners, GIC, ICONIQ, and XN. The round folds in $15 billion from hyperscalers including a $5 billion commitment from Amazon, plus infrastructure partners Micron, Samsung, and SK hynix. CFO Krishna Rao said the capital will fund safety and interpretability research, expand compute to meet demand, and scale Claude Code and Cowork. Source

2. Claude Opus 4.8 Ships With Better Judgment and Cheaper Fast Mode

Anthropic. Anthropic released Claude Opus 4.8 at the same list price as Opus 4.7, pitching it as a meaningful step up in honesty and judgment on agentic work, with the model asking clarifying questions and flagging its own uncertainties rather than guessing. Anthropic says it is roughly four times less likely than Opus 4.7 to let code flaws slip through, was the only model to finish every case on the Super-Agent benchmark, and scored 84% on the Online-Mind2Web browser-agent evaluation. The model is live via the API as claude-opus-4-8 at $5 per million input tokens and $25 per million output tokens, with a fast mode at $10/$50 that Anthropic frames as roughly three times cheaper than prior fast tiers. Source

3. Dynamic Workflows Bring Parallel Subagent Orchestration to Claude Code

Anthropic. Claude Code gained dynamic workflows, which let Claude write orchestration scripts that fan out tens to hundreds of parallel subagents in a single session and verify the results before surfacing them. The feature targets codebase-wide bug hunts, security audits, and large-scale migrations, running independent attempts and adversarial review until answers converge, with progress saved so interrupted jobs resume rather than restart. It is in research preview across the Claude Code CLI, Desktop, and VS Code extension for Max, Team, and Enterprise plans, with admin enablement required for Enterprise. Source

4. Anthropic Opens a Milan Office, Its Sixth in Europe

Anthropic. Anthropic opened an office in Milan, its sixth European location after London, Dublin, Paris, Zurich, and Munich, led by Head of Southern Europe Thomas Remy under MD International Chris Ciauri. The company cited existing Italian customers including Generali Group, Unipol Group, Angelini Pharma, Bracco Group, Enel Group, and Pirelli, and pointed to JAKALA deploying Claude across 3,000 seats and Satispay compressing an 18-month roadmap into seven months. The opening continues Anthropic’s geographic expansion that recently added Korea to its footprint. Source

5. Anthropic Hosts a Webinar on Putting Claude to Work for Security Defenders

Anthropic. Anthropic scheduled a webinar, “Claude Security: Putting Claude to Work for Defenders,” walking through how security teams can apply Claude to defensive workflows. The session lands alongside Anthropic’s recent enterprise security content on LLM-assisted source-code review and zero-trust postures for agents, signaling a sustained push to position Claude inside security operations. Source

6. Claude Code v2.1.153 and v2.1.154 Add Opus 4.8 Support and In-Agent Shell Execution

Anthropic. Claude Code tagged v2.1.153 and v2.1.154 yesterday, adding Opus 4.8 support with “high effort” as the default, dynamic workflows for cross-agent coordination, and Opus 4.8 fast-mode pricing. The releases also introduce shell command execution via ! <command> syntax inside agents, browser selection in the Chrome integration, a skipLfs option for plugin marketplace sources, and COLUMNS/LINES environment variables for status-line commands. The CHANGELOG remains the authoritative reference for teams pinning specific versions. Source