In a landmark moment for AI regulation, Anthropic has restored global access to its 'Fable 5' model following a multi-week shutdown triggered by a U.S. Commerce Department export-control directive. The model, along with Mythos 5, was pulled offline in mid-June after reports of a critical cybersecurity safeguard bypass. While Fable 5 is now back with enhanced classifiers, Mythos 5 remains restricted to vetted U.S. organizations. This event, coupled with OpenAI's recent staggered GPT-5.6 release, marks the first time the U.S. government has directly intervened in the commercial launch of frontier AI models.
The first half of 2026 has seen a decisive shift from chatbots to autonomous agents. Major labs including OpenAI (GPT-5.5) and Anthropic are pivoting toward 'agentic' systems capable of executing multi-step workflows across professional software. Success rates for these agents on real-world tasks have surged from roughly 20% in 2025 to over 77% as of mid-2026, leading to widespread enterprise adoption of persistent 'AI coworkers'.
As of July 2026, global spending on artificial intelligence has surpassed $2 trillion, driven by a transition from experimental pilots to massive enterprise-scale execution.
Which Nobel laureate and AlphaFold creator recently left Google DeepMind to join Anthropic?