Superintelligence does not exist yet, but Bloomington, Indiana, has already written it into a ban. On 4 November, the Bloomington Common Council voted 7–0 to adopt Ordinance 2026-29, adding Chapter 14.25 to the municipal code. It bars the distribution or deployment of superintelligence to any home or business in the city — including remote use. It targets developers and deployers; ordinary residents' everyday use is not covered, and it takes effect only after mayoral approval and publication.
What exactly does the ban cover?
The ordinance defines "superintelligence" as a system capable of significantly outperforming humans on essentially all cognitive tasks and able to operate without continuous human oversight — in other words, recursively self-improving AGI. What it bans is "distribution or deployment," drawn very broadly: using such a system through a browser, app, API or connected device inside the city counts as deployment, no matter where the servers are. According to the 4 November meeting record, an ordinary resident's personal use is not distribution or deployment; the ban aims at whoever supplies the system to others. Violators face up to $2,500 per day, and the city may also seek injunctions and refer cases to the Monroe County prosecutor. The ordinance also calls itself "interim": once the Indiana General Assembly enacts a comprehensive framework for "provably safe superintelligence," it expires 91 days after that framework takes effect.
Why would a small city move first?
Sponsor Dave Rollo calls it an "interim public safety perimeter" while state and federal lawmakers have not acted. The memo says the catastrophic consequences experts warn of make the risk "simply unacceptable." His sights go beyond Bloomington: if roughly 350 US cities of similar size or larger followed suit, frontier AI companies would face a massive compliance burden, and backers such as the Safe AI Forever Institute will keep lobbying in other cities. US states have already moved on AI legislation — California's AB 1883 and New York's chatbot law are two examples — and Bloomington has pushed lawmaking down to the municipal level. The original draft's exit conditions were extremely strict: a scientific consensus of 99.9999% safety, plus approval from more than 95% of registered voters. But according to Reason on 10 November, citing the meeting record, the 95%-of-voters clause was removed by amendment before the vote, and the final signed text had not been published as of 9 November.
What developers should watch; why ordinary users need not worry
For frontier AI companies, the real signal is not Bloomington itself, but the chance this gets copied: server location is no longer a shield — use inside the city counts as deployment — and if municipal rules spread across dozens or hundreds of cities, compliance costs will multiply by the number of cities. Three things are worth watching: whether the mayor signs within 20 days, whether any specific model is designated once the ordinance takes effect, and whether other cities follow. Critics have already named their legal weapons: Adam Thierer of the R Street Institute warns that hundreds of overlapping municipal rules would be a "compliance hell," and the ordinance could face Dormant Commerce Clause and First Amendment challenges; as of 10 November, no lawsuit had been filed. For ordinary users, the conclusion is much simpler: no model has been designated as superintelligence, and Indiana University students still use ChatGPT as before. Superintelligence itself does not exist yet — this ordinance bans something that has not arrived.
