A look at a proposed Massachusetts bill letting police use face recognition for image matching but not face surveillance, which could set the standard in the US
A Massachusetts bill restricting police use could set the standard for how the technology is regulated in America. Mastodon: @aregenberg@mastodon.online Mastodon: Alan Regenberg / @aregenberg@mastodon.online : Face recognition in the US is about to meet one of its biggest tests | MIT Technology Review https://www.technologyreview.com/ ...
Context & Ripple Effects
Massachusetts has been the strictest face-recognition jurisdiction in the country: after an ACLU-led campaign by Kade Crockford, lawmakers wrote a blanket ban on facial recognition by police departments and public agencies into the December 2020 police reform bill, and Boston had already enacted its own municipal ban that summer alongside San Francisco and Oakland.
The bill now under consideration reverses course at the edges — permitting state police to run image matching while still barring face surveillance and emotion recognition — which makes Massachusetts a test case for whether a state that banned the technology outright will legalize a narrow version, and what template that offers the rest of the country.
First-order effects
- State police gain a legally sanctioned image-matching capability they have lacked since the 2020 ban, while face surveillance and emotion recognition remain off-limits under the same statute.
- The ACLU and activists who won the original ban now face a narrower fight: defending the surveillance and emotion-recognition prohibitions rather than opposing any police use outright.
Second-order effects
- Cities like Boston, whose local ordinances were written for a total-ban environment, would sit inside a state framework that legalizes one narrow use — forcing municipalities to decide whether local bans still make sense.
- Vendors selling face-recognition matching tools gain a defined, regulated market in Massachusetts, and other state legislatures weighing bans versus rules get a middle-path model between Maine's near-total prohibition and unrestricted use.
Third-order effects
- If the pattern holds, US regulation converges on category-by-category carve-outs — matching allowed, surveillance and emotion recognition banned — replacing the city-by-city patchwork of outright bans documented since 2019, with Maine's stricter law as the ceiling and Massachusetts-style compromise as the likely floor.
- Explicitly excluding emotion recognition from permissible police use could establish the first standing legislative precedent against affect-recognition tools, a boundary other surveillance technologies may be judged against.
The trend: US states are shifting from blanket bans on face recognition toward narrowly scoped permitted-use statutes, with Massachusetts' match-but-don't-surveil bill as the next data point after Maine's stricter 2021 law.