The Fight Inside MAGA

While mainstream media was busy framing AI policy as 'Big Tech vs. Democrats,' the real war was happening inside MAGA itself.

According to CNN, in November 2025, David Sacks — Trump's AI czar — sat down with Mike Davis, a longtime Trump legal adviser, in Vice President JD Vance's office. It was not a friendly meeting.

Davis accused Sacks of trying to steamroll Congress and deploy AI on the country without adequate safeguards. Sacks said he was just executing Trump's orders. Both men walked in with sharp elbows.

Vance told them to figure it out together. They did — barely.

What the Truce Produced

A few weeks after that meeting, Trump signed an executive order designed to block states from enforcing their own AI regulations and directed the administration to build a 'single national framework.' CNN reported the order was shaped in part by both Sacks and Davis.

That order is widely expected to face legal challenges.

Meanwhile, the Center for AI Standards and Innovation — CAISI, housed inside the Department of Commerce — announced new pre-deployment testing agreements with Google DeepMind, Microsoft, and Elon Musk's xAI, according to CNBC. CAISI will 'conduct pre-deployment evaluations and targeted research to better assess frontier AI capabilities and advance the state of AI security.'

This builds on CAISI's 2024 deals with OpenAI and Anthropic. Those earlier agreements were renegotiated to reflect directives from Commerce Secretary Howard Lutnick and America's AI Action Plan.

What Changed Recently

The White House is now weighing a brand-new AI working group — separate from CAISI — that would include tech executives and government officials specifically tasked with developing pre-deployment oversight procedures, according to CNBC. The group may be established through an executive order. The White House told CNBC that any speculation about specific executive orders is just that — speculation — and policy announcements will come from Trump directly.

More than 60 MAGA-aligned allies have now formally told Trump he needs to vet AI before it goes public, according to Axios. The coalition includes working-class conservatives worried about job displacement, cultural conservatives focused on child safety, and longtime MAGA figures who simply don't trust Silicon Valley — regardless of how many checks tech CEOs wrote to Trump's inaugural fund.

The Anthropic Moment

The timing matters. CNBC confirmed that Anthropic's announcement of a powerful new frontier model caught government officials off guard last month. The model reportedly excels at capabilities that drew enough attention inside the administration to accelerate the working group discussions.

When a private company can drop a frontier model that surprises the people supposedly overseeing AI policy, officials scrambled to respond. The Sacks camp says speed is the point. The Davis camp says that's exactly the attitude that gets people hurt.

The Real Story in the Reporting

CNN's framing — 'Trump's AI push exposes a divide in the MAGA movement' — is technically accurate but misses the bigger picture.

This isn't just a MAGA internal squabble. This is a genuine policy debate about whether the U.S. government has ANY real leverage over the most powerful technology being built in human history — and whether voluntary testing agreements with the companies building that technology are worth the paper they're printed on.

The CAISI deals are voluntary. Google, Microsoft, and xAI are agreeing to be tested. They are NOT legally required to comply with findings. CAISI cannot force a company to delay a product launch. That gap — between 'we evaluated it' and 'we can stop it' — is where the entire framework falls apart.

Neither CNN nor CNBC spelled that out clearly. The 60+ MAGA signatories appear to understand it.

Lobbying Runs in the Background

While this internal debate plays out, tech companies aren't sitting still. According to CNN, companies have hired hundreds of lobbyists, donated millions to congressional campaigns, and are stockpiling cash in AI-friendly super PACs ahead of the next legislative push.

On one side: a well-funded industry with enormous influence over the politicians who are supposed to oversee it. On the other: a loose coalition of conservatives who don't trust those companies and are trying to slow things down without a clear legislative vehicle to do it.

Congress is the next front. And right now, Congress is outgunned.

For Americans

If you're an American worker, the question isn't whether AI is coming — it is. The question is whether anyone in Washington can require real safety standards before these systems reshape your industry, your kids' schools, and your daily life.

Right now, the answer is: we have voluntary testing deals and a working group that may or may not exist by executive order.