Data-center backlash after midterms pushes OpenAI to drop NDAs

This week’s midterm elections delivered Democrats a historic win that carried a clear anti-data-center message: dozens of local ballot measures imposing restrictions or moratoriums on data centers passed, discussion on TechCrunch’s Equity podcast showed. Panelists described the backlash as reaching across party lines rather than being confined to one side. In Texas, governor-elect Gina Hinojosa, who campaigned on a promise to restrict data centers in the state, defeated incumbent Greg Abbott.

The outcome lands awkwardly for Elon Musk, whose companies are now headquartered in Texas. Musk donated half a billion dollars — possibly more once final totals are counted — to Republican candidates this cycle, but, as TechCrunch’s Sean O’Kane put it on the podcast, he “seems to have basically nothing” to show for it, as his home state turns more Democratic than it has in a long time.

Panelists also linked the vote to how hyperscalers have handled data-center approvals — often requiring NDAs from elected or appointed officials and withholding details about a project’s scale in environmentally sensitive, water-scarce areas — and to a separate shift already underway on Wall Street, where investors have grown reluctant to keep rewarding hyperscalers and AI companies for continued heavy spending on data-center buildouts. TechCrunch’s Kirsten Korosec argued that even early signals of tighter rules can move corporate behavior quickly, pointing to OpenAI’s abrupt shift as a case in point.

In apparent response, OpenAI has published what it calls a “compute campus policy blueprint,” under which the company says it will stop requiring NDAs, share more about its plans, and supposedly operate with less secrecy — rebranding its data centers as “compute campuses” in the process. Panelists were skeptical the renaming would shift public sentiment, with one dismissing the attempt as unlikely to change the underlying political pressure.

Why it matters

The episode suggests other hyperscalers now face the choice OpenAI just made: trade some secrecy for public buy-in before more ballot measures or state leaders like Hinojosa make that call for them. It also signals that AI infrastructure spending, long treated as untouchable, is now exposed to both electoral and investor pushback at once.