Anthropic released Claude Fable 5 on June 9, making its most powerful Mythos-class AI available to the public for the first time, with hard blocks on cybersecurity and biology queries and a new mandatory 30-day data retention policy for all enterprise traffic.
Microsoft unveiled MAI-Thinking-1 at Build 2026, its first reasoning model built entirely in-house without OpenAI or Anthropic distillation. With 35 billion parameters and a 256,000-token context window, it matches Claude Opus 4.6 performance — signaling Microsoft’s push for AI independence.
OpenAI has released an updated GPT-Rosalind life sciences model combining GPT-5.5 capabilities with drug discovery intelligence. The model uses 31 percent fewer tokens while improving accuracy, with Novo Nordisk receiving early access alongside Amgen, Moderna, and other pharma partners.
Apollo Global Management has built a formal AI disruption risk framework to screen every software investment it considers, dividing software into 12 to 14 categories ranked by susceptibility to AI. The firm is redirecting toward HALO assets as private equity’s massive software bets come under scrutiny.
Figure AI has hit a production rate of one Figure 03 humanoid robot per hour at its BotQ facility, a 24-times improvement in 120 days. The company can now produce 12,000 units per year, while first-pass yields exceed 80 percent.
Amazon has officially deployed over one million robots across its global fulfillment operations while unveiling a new AI-powered Proteus robot that responds to plain language commands. The DeepFleet AI system coordinates the fleet with a ten percent efficiency improvement.
The American tech industry has now cut 183,966 workers in 2026, with AI becoming the single most cited reason for job cuts for the first time. Fifty-five percent of all layoff announcements this year have explicitly named AI, automation, or machine learning.
Eight of Japan’s largest financial institutions have joined a new AI collaboration between NEC and Anthropic, deploying Claude across banking and insurance operations even as regulators warn that Mythos could be weaponized for cyberattacks.
China has announced plans to spend 2 trillion yuan — roughly $295 billion — over the next five years building a nationwide network of interconnected AI data centers, requiring 80 percent domestic chips and effectively locking out Nvidia.
OpenAI signed a five-year, $300 billion contract with Oracle to build 4.5 gigawatts of Stargate data center capacity in the US, the largest data center infrastructure deal ever recorded. The contract converts the Stargate joint venture announced in January into a binding commercial agreement.