THIS WEEK IN AI
Candice Bryant Consulting
Strategic Intelligence & Public Affairs
Each week I synthesize the key developments in AI and public policy so you don't have to. Here's what I'm watching:
• President Trump announced Nvidia will be allowed to sell H200 chips to China with a 25% cut for the U.S. government
• The administration signed an executive order challenging state AI regulations, prompting pushback from state lawmakers
•The federal government launched U.S. Tech Force to hire more engineers and DOD launched GenAI.mil
Taken together, this week's developments underscore critical priorities for tech companies: addressing national security concerns from Congress, navigating competing federal and state regulatory priorities, and deepening government partnerships.
How to Read This Newsletter: This Week in AI applies the same principles of concise writing used to brief busy policymakers. Each entry contains a summary and analysis—if you already know what happened, jump straight to "What I'm Watching" for insights.
1. U.S. Allows Nvidia to Sell H200 Chips to China
President Trump announced Nvidia will be allowed to sell H200 chips to approved customers in China, with the U.S. government receiving 25 percent of revenue. Trump indicated the same framework will apply to AMD, Intel, and other U.S. chip companies. Nvidia's most advanced Blackwell chips remain restricted, but the H200 is about six times more powerful than the H20 chip previously approved for export. Press reports indicate the the H200's 18-month lag behind current Blackwell chips was part of the rationale for the decision. The Department of Commerce is finalizing implementation details.
Critics argue the move undermines the only effective policy tool for slowing China's AI progress and say it remains unclear if China will accept the concession.
What I'm watching:
This decision has faced scrutiny on both sides of the aisle and Sen. Elizabeth Warren has called for Nvidia CEO Jensen Huang and Commerce Secretary Howard Lutnick to testify before Congress. Chip companies should be prepared to answer tough questions about whether this creates a slippery slope where increasingly powerful chips become exportable over time and how that impacts national security. Companies will also need to be ready to explain the 25 percent revenue-sharing arrangement to investors and integrate it into China market forecasts.
2. Trump Signs Executive Order Challenging State AI Laws
As anticipated, President Trump signed an executive order challenging state AI regulations, arguing that a patchwork of state laws harms U.S. competitiveness. The order directs the Justice Department to establish a task force to challenge state AI laws, instructs the Commerce Department to identify regulations it deems problematic, and signals potential restrictions on certain federal funding, including broadband grants.
What I'm watching:
With preemption having failed in Congress and the executive branch lacking authority to ban state AI laws outright, this executive order shifts the fight to a new arena. Implementation will unfold through funding decisions and court challenges. Thirty-eight states have enacted AI measures this year and numerous state lawmakers have prefiled legislation for 2026 and vowed to continue moving forward despite the executive order. In the meantime, companies will need to navigate compliance with state laws that remain on the books but face federal challenge, creating a prolonged period of regulatory uncertainty.
3. DOD Launches AI Platform as Federal Government Expands Tech Hiring
The federal government is continuing efforts to integrate AI into government operations, pairing new internal tools with targeted hiring.
The Department of Defense announced GenAI.mil, an internal AI platform powered by Google's Gemini models. The platform is limited to unclassified work and can be used to conduct research, format documents, and analyze data. Separately, the administration launched the "U.S. Tech Force," a campaign to hire 1,000 engineers—including AI specialists—for two-year roles embedded across federal agencies. The initiative builds on a similar Biden-era program.
What I'm watching:
These moves highlight strategic opportunities to deepen public-private partnerships. Major tech companies have agreed to consider Tech Force alumni for future roles, creating a pipeline of technical talent with direct government experience and understanding of federal AI priorities. The GenAI.mil launch could serve as a model for other companies to develop purpose-built platforms for government customers. Both developments underscore the importance of tech companies positioning themselves as long-term government AI partners rather than just vendors.
