OpenAI's Political Playbook Is Already Being Written for 2028

NewsOnScale Staff

November 5, 2026

There's a version of this story where a technology company engaging with the political process is entirely normal. Companies hire lobbyists. They fund PACs. They send executives to testify before Congress. That's how the system works, and OpenAI is hardly the first Silicon Valley firm to figure that out.

But OpenAI is not a normal technology company, and 2028 is not a normal election cycle to be preparing for.

## Why This Moment Is Different

What separates OpenAI's early positioning from standard corporate politicking is the nature of the product itself. OpenAI doesn't make software that helps businesses run payroll or stream video. It builds systems that generate text, synthesize information, hold conversations, and increasingly, take autonomous action on behalf of users. Those systems are already being used to draft political messaging, conduct opposition research, and potentially — depending on how guardrails evolve — to operate at scale inside the information environments where voters form opinions.

A company that profits from those capabilities working simultaneously to shape the laws governing those capabilities is a structural conflict of interest that deserves more scrutiny than it typically receives.

To be clear: there is nothing inherently illegal about what OpenAI is reportedly doing. Early political engagement is legal. It is also strategic. The 2026 midterms will determine committee compositions and regulatory appetites in Congress. The 2028 presidential race will likely feature AI as a central policy battleground — both as a campaign tool and as a subject of proposed legislation. Getting ahead of that environment, shaping the vocabulary lawmakers use, and building relationships with candidates before they're frontrunners is exactly what well-resourced organizations do.

The problem is that most well-resourced organizations don't also control the infrastructure those same candidates might use to communicate with voters.

## The Agent Economy Dimension

This story sits squarely at the intersection of two of the most consequential trends in the AI economy: the rapid deployment of autonomous AI agents and the growing financial footprint of AI-aligned political actors.

As we've covered previously, AI PACs spent heavily during the last cycle with mixed results. The lesson those groups are likely drawing is not to spend less, but to spend earlier and smarter. If OpenAI is already in conversations — formal or informal — about the 2028 landscape, it is positioning itself to be the reference point from which AI policy gets debated rather than a subject of that debate.

That's a meaningful distinction. There's a difference between a company that defends its interests in a regulatory hearing and a company that helped write the framing documents the hearing is built around.

## What Accountability Looks Like Here

The accountability questions that follow from this are specific and answerable with the right disclosure infrastructure. Which candidates or campaigns has OpenAI engaged with, and in what capacity? Are those conversations happening through registered lobbyists — which creates a paper trail — or through informal channels, speaking invitations, and research partnerships that don't? Has OpenAI offered access to its tools, its models, or its infrastructure to any political committee or affiliated organization?

None of these questions are accusations. They are the basic informational requirements for the public to evaluate whether a company with this much power over communication technology is operating inside or outside appropriate boundaries.

## The Longer Arc

The United States has spent decades figuring out how to regulate the relationship between media companies, telecommunications companies, and political campaigns. That process was slow, imperfect, and often captured by the very industries it was meant to govern. AI presents the same challenge at faster speed and higher stakes.

OpenAI beginning its 2028 preparations now is, in one sense, just competent long-term planning. In another sense, it's a signal that the window for establishing meaningful guardrails around AI's role in democratic processes is closing faster than policymakers seem to recognize.

NewsOnScale will be tracking this thread through every development. The public deserves a clear accounting of who is shaping AI policy, what they stand to gain, and whose interests are going unrepresented in those rooms.