The governor of Colorado has signed an AI bill passed at the very end of the state's legislative session for 2024. The new law will take effect in 2026. The Colorado law prohibits “algorithmic discrimination” in the use of “high-risk” AI systems used to make “consequential decisions.” The bill also provides a definition of an AI system (a machine-based system that infers from inputs how to generate outputs).
Why It Matters
The backers of the Colorado bill worked with legislators in Connecticut to draft this legislation; Connecticut introduced a similar bill but failed to pass it. The fact that we now have one general AI law on the books, however, and the cooperation across state lines to “standardize” legislative thinking about AI legislation, foretell a busy legislative season next spring, when we can probably expect to see copycat bills in other state legislatures. This bill is also notable because, even though he signed it into law, the governor has expressly invited the legislature to revisit the bill and broaden it before it takes effect. Finally, the bill has given us some useful definitions that can be incorporated into all our thinking about AI systems and how we evaluate them – what is an AI, what are high-risk AIs, and what are consequential decisions. Having this law on the record now presents some “best practices” even for companies that are not subject to the law, in other words.
Subscribe to Taylor English Insights by topic here.