HireQwikMassachusetts Never Passed an AI Hiring Law — It Just Said Your Existing Liability Already Covers It
Massachusetts Never Passed an AI Hiring Law — It Just Said Your Existing Liability Already Covers It
Massachusetts has no AI-specific hiring statute in 2026. What it has instead is an advisory: in April 2024, Attorney General Andrea Campbell told developers, suppliers, and users of AI systems that Massachusetts’ existing consumer protection, anti-discrimination, and data security laws already apply to AI in full, including in hiring — no new bill required. For an Indian GCC or BPO screening candidates into Massachusetts-based roles, that’s easy to miss, because there’s no bill number to bookmark, no committee hearing to track, no signing date on a calendar. That’s exactly why it’s worth tracking.
No new law, same exposure
The advisory doesn’t create new obligations so much as it removes the argument that old ones don’t reach AI. It states plainly that Massachusetts’ anti-discrimination law prohibits deploying AI that discriminates on a legally protected characteristic — including algorithmic decision-making that relies on discriminatory inputs and produces discriminatory results — the same as it would prohibit a human doing it. It also pulls in Chapter 93A consumer protection law and the state’s data security “Standards” under Chapter 93H, meaning an AI vendor’s claims about its own product become fair game for a false-advertising complaint, not just its hiring outcomes.
That’s a wider net than a narrow hiring-specific statute would be, and it’s also harder to spot on a compliance checklist, because it doesn’t announce itself the way a named act does.
Where this differs from the four-state pattern
We’ve written before about how California, Illinois, Texas, and Colorado each built a different legal theory for AI hiring liability after the EEOC pulled its federal guidance — vendor-liability disparate impact, a private right of action, an intent-only standard, and a reasonable-care affirmative defense, respectively. Massachusetts is a fifth pattern, and it’s the one most likely to get missed by a team that’s literally searching state legislatures for “AI hiring bill” to build their tracking list. There isn’t one to find. The exposure sits inside statutes that have existed for decades, reinterpreted for a technology those statutes never mentioned by name.
The practical risk of missing it isn’t abstract. A compliance checklist built around named AI statutes will flag California and Illinois and skip Massachusetts entirely, right up until a candidate complaint or an AG inquiry arrives citing Chapter 93A instead of anything that looks like an “AI law.”
What the advisory actually flags for a vendor
The consumer-protection angle is the sharper edge here, and it’s aimed at vendors as much as employers. The advisory specifically calls out misrepresenting an AI system’s reliability, safety, or performance — including whether it’s “free from bias” — as a consumer-protection violation on its own, independent of whether a specific hiring decision goes wrong. A vendor that markets its screening tool as bias-free or a fixed accuracy percentage it can’t substantiate isn’t just making a marketing claim; under this advisory, it’s making a claim Massachusetts can act on directly.
That’s a standard worth holding yourself to even outside Massachusetts. It’s also why we’ve never claimed our own screening reduces bias to zero or hit a specific accuracy number we can’t stand behind — the honest version is a measured outcome (an 89% cut in HR screening time in one pilot, verified and specific) rather than a claim about the model’s fairness that no one has actually benchmarked.
What this means for an India-based team
If you’re screening candidates into Massachusetts roles, or selling screening software to a company that does, the advisory doesn’t require a new process built specifically for Massachusetts. It requires the same discipline the four-state pattern already demands: know why a candidate was rejected in terms you could explain to a regulator without pulling in an engineer, don’t let a vendor’s marketing claims outrun what’s actually been measured, and keep a record that survives scrutiny under whichever law — named statute or reinterpreted old one — eventually asks for it.
The contrarian read: “no AI hiring law” in a state should never be read as “less exposure.” Sometimes it means the opposite — the exposure was already sitting in statutes nobody thought to check, waiting for an attorney general to say so out loud.
If you’re evaluating what an audit-ready screening record looks like across states with and without a named AI statute, talk to us about what that actually requires.
See HireQwik in action
Book a 30-minute demo — bring a live JD and we'll screen your own candidates against it.