Microsoft has launched Copilot Cowork, a new enterprise AI agent built in collaboration with Anthropic, enabling autonomous multi-step task execution within Microsoft 365. The feature is part of the “Wave 3” rollout and will be available as a research preview in March, with Agent 365 orchestration platform reaching general availability on May 1.
On March 9, 2026, China’s National People’s Congress (NPC) announced plans to intensify research into AI-specific legislation as part of its 2026 legislative agenda, signaling a strategic shift toward formalizing AI governance.
Caitlin Kalinowski, who led OpenAI’s robotics and hardware efforts, resigned on March 7, 2026, protesting the company’s rushed agreement with the Department of Defense and lack of defined guardrails around surveillance and lethal autonomy.
OpenAI has begun rolling out GPT‑5.4, a major upgrade that integrates reasoning, coding, and agentic capabilities into a unified architecture. The model introduces native financial‑services plugins for Excel and Google Sheets and supports a 1‑million‑token context window—enabling it to process entire codebases or lengthy financial documents in a single pass.
Google DeepMind has expanded enterprise access to its advanced reasoning model, Gemini 3 Deep Think, while also launching the cost-efficient Gemini 3.1 Flash‑Lite API. These moves mark a significant push in enterprise AI capabilities and accessibility.
WisdomAI has launched a Federated Agentic Intelligence Platform that enables enterprises to deploy AI agents across data silos without moving data, marking a significant leap in autonomous enterprise analytics.
The Biden administration has directed all federal agencies to immediately cease using Anthropic’s AI technologies, citing escalating national security risks tied to the company’s supply chain.
JPMorgan Chase is increasing its 2026 technology budget by nearly $2 billion to $19.8 billion, with approximately $1.2 billion earmarked for AI-driven projects spanning customer service automation, personalized client insights, and engineering tools.
On March 5, 2026, OpenAI introduced a dedicated “Adoption” channel on its blog, aimed at providing C‑suite executives and enterprise leaders with practical frameworks, playbooks, and value models to translate AI capabilities into operational business outcomes.
Connecticut legislators introduced a sweeping AI policy package including an AI regulatory sandbox, an Artificial Intelligence Policy Office, and transparency and anti‑discrimination mandates, aiming to balance innovation with consumer protection.