Anthropic Launches Claude Fable 5 and Calls for Global AI Safety Oversight
Anthropic has released Claude Fable 5, the first of its 'Mythos-class' models available for general use, which set a new benchmark record of 80.3% on SWE-Bench Pro. Simultaneously, the company published two policy frameworks urging governments to implement stricter oversight for frontier models and economic safeguards to protect workers from AI-driven disruption.
The AI industry is shifting from reactive chatbots to autonomous 'Agentic Infrastructure.' Major developments this week, including Mastercard's 'Agent Pay' and Google's deployment of engineering agents, signal a move toward AI systems that can independently manage financial transactions, complex codebases, and industrial workflows.
News roundup
OpenAI and Anthropic have both filed confidentially for US initial public offerings, with OpenAI targeting a valuation of up to $1 trillion.
Mastercard launched 'Agent Pay for Machines,' a programmatic payment system allowing AI agents to conduct microtransactions at machine speed.
The Financial Stability Board (FSB) published a new framework outlining 12 'sound practices' for the responsible adoption of AI within the global financial sector.
Google Cloud announced a collaboration with Randstad Digital to deploy 'Forze Mirate,' an agentic AI designed to automate knowledge transfer for high-turnover engineering teams.
PTC introduced 'PTC Orbit,' a cloud-native asset intelligence solution that uses AI to unify product lifecycle data for manufacturers.
Did you know?
As of April 2026, Anthropic's annualized revenue reached approximately $30 billion, surpassing OpenAI's $25 billion for the first time, despite OpenAI maintaining a significantly larger consumer user base.
Quiz
Which new Anthropic model, released in June 2026, is the first 'Mythos-class' AI available for general public use?