OpenAI’s Lehane likens backing a data centre ban to the 2002 Iraq vote

OpenAI chief global affairs officer Chris Lehane says AI will be the leading issue in the 2028 US presidential election. His memo, “2028: The AI Election,” outlines five questions candidates should address in developing an AI agenda; Politico published the memo in full on Monday.

Analysis

The actionable signal is not that AI will dominate the 2028 campaign, but that a major AI developer is publicly framing policy as a central competitive variable. That raises the prospect of wider dispersion among AI firms: companies able to document safety controls, domestic investment, and compliance may gain procurement access and lower policy risk, while less-established developers could face disproportionate compliance costs. The same scrutiny could also constrain deployment or raise costs for data-center power, permitting, and model governance.

There is no near-term earnings catalyst in the information provided. Over the next 1–3 months, the useful signal will be whether candidates and policymakers translate broad debate into specific proposals on training-data rights, liability, export controls, energy, or federal procurement. Over 6–18 months, regulatory fragmentation and election-related uncertainty could delay investment decisions even before any law changes. The memo is an interested party’s policy framing, not evidence that voters rank AI first or that policy will move in a particular direction.

Contrarian point: election attention can increase policy uncertainty without producing restrictive outcomes; it may instead accelerate public-sector adoption and infrastructure spending. Avoid treating the headline as a directional signal for AI equities.

Key Decisions for Investors

• No immediate thematic position: the article supplies no candidate proposals, legislative text, or company-level exposure data to support a differentiated long or short.
• Monitor policy specifics rather than campaign rhetoric—especially liability, training-data rules, export controls, procurement standards, and data-center permitting. Reassess only when proposals have identifiable implementation paths.
• For AI-exposed holdings, track evidence of compliance cost, deployment delays, and government-contract eligibility; these are more likely to separate winners from losers than broad AI sentiment.
• Falsifier for the rising-policy-risk thesis: bipartisan proposals remain nonbinding or narrowly scoped while adoption and public procurement expand without material compliance or permitting delays.