Add v64otd.com to your daily reading list — the ledger doesn't lie.
June 6, 2026
On June 4 — three days ago — Representatives Jay Obernolte of California and Lori Trahan of Massachusetts released a 269-page discussion draft of legislation called the Great American Artificial Intelligence Act of 2026. It was co-sponsored by four additional members of Congress from both parties. It received almost no sustained mainstream press coverage — because the OpenAI IPO roadshow, the SpaceX $75 billion valuation, and the Iran deal negotiations were all competing for the same finite attention. The bill that will determine how the most powerful technology in human history is governed in the United States dropped on a Thursday afternoon and spent the weekend largely unread.
This dispatch reads it for you. And what it contains deserves the same scrutiny that Section 224 of the NDAA received — because the Great American AI Act will shape your data, your workplace, your children's education, your interactions with law enforcement, and the balance of power between technology companies and the people whose lives they are increasingly making decisions about. You funded the technology through your data and your consumer behavior. You deserve to know what Congress has decided to do with the power it creates.
What the Bill Actually Is — And What It Is Not
The Great American AI Act is a discussion draft — not yet formally introduced as legislation. A discussion draft is a pre-legislative document released for public and congressional comment before a bill is formally filed. This distinction matters: the bill can still be substantially changed before it becomes law, and the release of a discussion draft is an invitation for stakeholder input that typically precedes formal introduction by weeks or months. The public comment window that this draft represents is the best opportunity citizens have to influence what the final legislation looks like. Most people do not know it exists. That is the problem this dispatch is here to address.
The bill was authored primarily by Representative Jay Obernolte — a Republican from California who has a computer science degree and is the only member of Congress who is also a professional video game developer. His co-author, Lori Trahan, is a Democrat from Massachusetts with a technology policy focus. The bipartisan authorship is real — this is not a partisan document. It is also not, according to its most prominent critics, an adequate one.
What the Bill Contains — The Six Most Important Provisions
The Great American Artificial Intelligence Act of 2026 is bipartisan legislation to create a federal framework for how the United States governs artificial intelligence. Here are the six provisions that matter most to ordinary Americans: sec
Provision 1 — Federal preemption of state AI laws for three years. This is the most immediately consequential and most controversial element of the bill. The draft includes a three-year preemption of state laws related to AI development that has previously generated significant pushback. What this means in practice: for three years after the bill's passage, no state — not California, not Texas, not New York — can pass its own laws specifically regulating the development of AI models. The federal framework becomes the only framework. It would prohibit states and local governments from enacting or enforcing laws specifically regulating the development of artificial intelligence models while leaving oversight largely to a federal government that has repeatedly failed to pass meaningful AI protections. Brad Carson, President of Americans for Responsible Innovation, called this a "generational mistake" that could weaken existing protections for the public. Supporters say federal uniformity prevents a confusing patchwork of state rules that could slow innovation. What is not disputed is that three years is a long time in AI development — and that during those three years, whatever states have already built for their residents' protection gets suspended. secNPR
Provision 2 — Focus on frontier models only. The bill targets what it calls "advanced AI systems" — the most powerful frontier models like GPT-5, Claude, and Gemini. It does not apply to the vast majority of AI systems currently deployed in hiring, lending, healthcare, criminal justice, and education. The U.S. approach is narrower than the EU AI Act, which applies broad risk categories across many types of AI systems. Instead, this bill zeros in on the most powerful frontier models. The practical consequence: the AI system that screens your job application, determines your loan eligibility, or recommends your child's educational intervention is almost certainly not covered by this bill's protections. sec
Provision 3 — $100 million for a Center for AI Standards and Innovation. The discussion draft would authorize $100 million per fiscal year for a Center for AI Standards and Innovation and calls for oversight in government AI adoption. This is the bill's most straightforward provision — public investment in the technical infrastructure for AI safety standards. Its critics note that $100 million per year is a rounding error compared to what AI companies are spending on development — OpenAI alone is burning $14 billion per year — and that standards without enforcement are aspirations, not protections. Stocktitan
Provision 4 — Model safety and transparency requirements. The bill requires developers of covered frontier AI models to conduct safety testing before deployment, maintain documentation of training data and model capabilities, and report significant safety incidents to a federal authority. These are meaningful requirements. They are also largely what the largest AI companies are already doing voluntarily — or claim to be doing. The bill's critics from the left note that it provides no independent verification mechanism — the companies self-report and the government accepts those reports. The bill's critics from the right note that compliance costs will disadvantage smaller competitors and entrench the largest incumbents.
Provision 5 — Workforce impact assessment. The bill requires covered AI developers to assess and report on the workforce impacts of their models — specifically the displacement of workers whose jobs are automated by AI systems. This is a genuinely new requirement with no equivalent in current law. It does not require companies to prevent displacement or compensate displaced workers. It requires them to measure and report it. Whether a disclosure requirement without a remedy requirement is meaningful protection is a legitimate debate. What is not debatable is that this provision acknowledges, for the first time in federal legislation, that AI-driven workforce displacement is a policy concern the government has an obligation to track.
Provision 6 — Preemption of stronger state protections. This is separate from the three-year development preemption in Provision 1 and arguably more alarming. States could still regulate how AI is used in areas like privacy, employment, and consumer protection. However, the preemption of development regulation means that states that have already passed stronger AI safety requirements — Colorado, California, and Texas among them — may find those laws superseded by a federal framework that consumer advocates describe as substantially weaker. Public Citizen called it "a disastrous proposal that Big Tech is celebrating." CBS NewsCNN
What the Bill Does Not Contain — The Gaps That Matter
The Consumer Federation of America's response to the bill is worth quoting at length because it names specifically what the bill omits: "Americans are desperate for bold legislative action on AI that protects their data, pocketbooks, livelihoods, and communities. The so-called Great American AI Act fails to do any of this. Congress could move any number of strong and substantive bipartisan bills already introduced, but this effort would instead create a weak and convoluted regulatory state that lets AI companies continue to do whatever they want with impunity while in many cases preventing states from protecting their own residents from the harms they are actually experiencing." NPR
The CFA specifically identified what the bill does not contain: no meaningful privacy protections for the data used to train AI models, no prohibition on AI-generated scams, no liability framework for chatbot-caused harm, no restrictions on surveillance pricing — the practice of using AI to charge different customers different prices based on behavioral profiling — and no data center accountability provisions.
The data privacy omission is the most significant. The bill that governs the most data-intensive technology in history contains no meaningful restriction on how that data is collected, used, or monetized. Every conversation you have with an AI system — every query, every document you upload, every personal disclosure you make to a chatbot — contributes to the training data that makes those systems more valuable. That data is currently yours in theory and the AI company's in practice. The Great American AI Act, as drafted, does not change that relationship. It does not require consent for training data use. It does not establish data portability rights. It does not create a right of deletion. It does not restrict cross-platform data sharing. It leaves the data architecture exactly as it currently exists and asks the companies that benefit from it to report on safety incidents.
The Preemption Problem — Why This Matters for Texas Specifically
The three-year state preemption is not abstract for Texas. Texas passed its own AI legislation — the Texas Responsible AI Governance Act — in 2025, establishing requirements for high-risk AI systems used in consequential decisions affecting Texas residents. That legislation was the result of Texas legislators responding to documented harms: AI systems used in hiring that showed documented racial bias, AI-driven insurance pricing that disadvantaged rural Texans, and facial recognition deployments by Texas law enforcement that produced documented false positives disproportionately affecting Black residents.
If the Great American AI Act passes as drafted, those Texas protections — enacted by the Texas Legislature, signed by the Texas Governor, in response to documented harms to Texas residents — would be superseded by a federal framework that consumer advocates describe as substantially weaker. The same dynamic applies to every state that has moved faster than Congress on AI protection. The bill does not preempt stronger protections because the federal standard is stronger. It preempts them because Big Tech prefers federal regulation it helped design to state regulation it did not.
The V64OTD dispatch on Section 224 documented how a single sentence buried in a defense authorization bill proposed to fuse the US and Israeli militaries without public debate. The Great American AI Act is 269 pages — substantially more text — and it lands on a Thursday afternoon to almost no coverage. The mechanism is different. The dynamic is the same: consequential legislation, minimal public awareness, maximum Big Tech satisfaction.
What Happens Next — And Why the Next 60 Days Matter
The discussion draft status of the bill is the most important fact about it for anyone who wants to influence what it becomes. Discussion drafts are released specifically to invite input before formal introduction. The sponsors — Obernolte and Trahan — have explicitly stated they want stakeholder feedback before finalizing the bill's text. That feedback window is open right now and will likely close within 60 days as the bill moves toward formal introduction ahead of the fall legislative calendar.
The provisions most susceptible to change based on public pressure are the state preemption scope and the data privacy omissions — because these are the areas where the bill has drawn the sharpest bipartisan criticism. Republican members from states with existing AI legislation have privacy and sovereignty objections to federal preemption. Democratic members have consumer protection objections to the data privacy gaps. Both sets of objections map to the same legislative fix: narrow the preemption scope and add meaningful data rights provisions. That fix is achievable — but only if the people who would benefit from it make enough noise in the next 60 days.
Contact your representative and two senators specifically about the Great American AI Act discussion draft. Ask them three specific questions: Does the three-year state preemption supersede Texas's existing AI protections? Why does a bill governing the most data-intensive technology in history contain no meaningful data privacy provisions? Who specifically consulted with Big Tech in drafting this bill and what did those consultations produce? Those questions go to the record. Make them answer in writing.
Read the bill yourself. The full 269-page discussion draft is publicly available at obernolte.house.gov. You do not need to read all 269 pages — read the definitions section, the preemption provision, and the enforcement mechanism. Those three sections tell you most of what you need to know about whose interests this bill was designed to serve.
The AI era is being governed right now — in 269 pages that most Americans will never read, during a news cycle dominated by an IPO roadshow and a ceasefire negotiation. The ledger is open. Whether anyone reads it before it becomes law is, as always, the question.
V64OTD // 269 PAGES. BIG TECH IS CELEBRATING. DO YOU KNOW WHY?