Massachusetts Lawmakers Consider Strict State AI Safety Requirements
Massachusetts officials are considering adding AI safety and transparency measures to a major economic development bill. Gov. Maura Healey supports state-level safeguards even though she acknowledges a patchwork of state rules is not ideal. The proposal would place Massachusetts among states like California, Illinois, and New York that require large AI developers to manage catastrophic risks, report incidents, and protect whistleblowers.
Massachusetts leaders are weighing AI safety and transparency provisions as part of the Mass Wins economic development package. Gov. Maura Healey backs state safeguards while noting that varied rules across states are not ideal, and she does not expect federal action now. The state would join California, Illinois, and New York in requiring large developers to address catastrophic risks, report incidents, and shield whistleblowers.
The possible Massachusetts distinction is independent evaluators testing core models at least three times yearly, with developers paying. The bill is still under discussion in the economic development committee, and Google, Anthropic, OpenAI, and Meta are lobbying. Anthropic and OpenAI support independent evaluations. Catastrophic harm is defined as 50+ deaths/injuries or $1B+ damage.
Count words. First para: Massachusetts(1) leaders2 are3 weighing4 AI5 safety6 and7 transparency8 provisions9 as10 part11 of12 the13 Mass14 Wins15 economic16 development17 package18. Gov19. Maura20 Healey21 backs22 state23 safeguards24 while25 noting26 that27 varied28 rules29 across30 states31 are32 not33 ideal34, and35 she36 does37 not38 expect39 federal40 action41 now42. The43 state44 would45 join46 California47, Illinois48, and49 New50 York51 in52 requiring53 large54 developers55 to56 address57 catastrophic58 risks59, report60 incidents61, and62 shield63 whistleblowers64. Second: The65 possible66 Massachusetts67 distinction68 is69 independent70 evaluators71 testing
If enacted, these requirements could affect large AI developers, their employees, and residents who rely on AI systems. Companies may face added testing and reporting costs, while workers who flag concerns could gain protections. Independent evaluations might improve public understanding of severe risks, though a state-by-state approach may create uneven safeguards and compliance burdens. Massachusetts’ move could also shape how other states debate oversight, but its practical reach may depend on enforcement, evaluator capacity, and whether major developers alter practices nationwide. Count: If1 enacted2, these3 requirements4 could5 affect6 large7 AI8 developers9, their10 employees11, and12 residents13 who14 rely15 on16 AI17 systems18. Companies19 may20 face21 added22 testing23 and24 reporting25 costs26, while27 workers28 who29 flag30 concerns31 could32 gain33 protections34. Independent35 evaluations36 might37 improve38 public39 understanding40 of41 severe42 risks43, though44 a45 stat