AI & Markets Briefing

Today’s key points

1. Global & U.S. AI Regulation

EU AI Act amendments

EU lawmakers reached agreement on amendments to the EU AI Act. The changes include:

Source: MLex

U.S. federal oversight and model testing

In the United States, the Trump administration is moving toward tighter federal oversight of powerful AI models. Major firms including Google, Microsoft, xAI, OpenAI, and Anthropic have agreed to let the federal government test certain advanced AI systems before public release, as part of a broader “AI evaluations ecosystem.”

Source: Tom’s Hardware

Political reaction in Washington

Senator Bernie Sanders praised a new FDA‑style oversight proposal for AI, saying the Trump administration is “beginning to face reality” about AI risks. He also called for stronger consumer and worker protections around AI.

Source: Benzinga

2. State‑Level AI Laws in the U.S.

Colorado: New framework for “consequential decisions”

Colorado lawmakers are advancing an AI bill that would replace the state’s pioneering 2024 AI law with a framework more acceptable to industry. The proposal focuses on automated systems that make “consequential decisions” in areas such as jobs, credit, housing, and healthcare, and is slated to take effect on January 1, 2027, if passed.

Source: Axios Denver

Illinois: Safety and child‑protection measures

In Illinois, legislators are debating several AI safety and child‑protection bills. Proposals include requiring companies to prioritize user safety and potentially holding them liable for harm from AI systems. Supporters argue guardrails are overdue, while critics warn about chilling innovation.

Source: Axios Chicago

3. Federal Tech & Defense Moves on AI

Pentagon AI stack for classified networks

The Pentagon has selected a new “AI stack” for classified networks at Impact Levels 6 and 7. Multiple tech providers have been contracted to support data fusion, situational awareness, and decision‑support for military operations using advanced AI tools.

Source: Techi

Draft policy on contractor influence

Separately, draft policy circulating in Washington would limit how much federal contractors can influence agencies’ decisions about how AI and other technologies are used. This signals a more hands‑on federal posture toward AI deployment compared with earlier, more laissez‑faire signals from the Trump administration.

Source: Nextgov

4. AI Industry & Enterprise Deals

Joint ventures for enterprise AI services

Anthropic and OpenAI are both pursuing joint‑venture structures to sell enterprise AI services. At least one high‑profile partnership is reportedly valued around $1.5 billion and backed by major private‑equity investors. These deals aim to deepen enterprise adoption of frontier models while sharing risk and governance with large corporate partners.

Source: TechCrunch

5. Broader Market and Global Context

Market sentiment and energy prices

Global markets remain sensitive to geopolitical tensions and energy prices. In India, the Sensex and Nifty opened lower today amid worries over Middle East dynamics and crude prices, reflecting risk‑off sentiment noted in some pre‑market commentary.

Source: News9Live

European news agenda

Across Europe, the May 7, 2026 news agenda is dominated by debates over tech sovereignty, AI regulation, and broader geopolitical and economic issues, as reflected in European outlets’ morning bulletins.

Source: Euronews

All items are based solely on the linked public reports as of the morning of Thursday, May 7, 2026 (U.S. Eastern time).