Colorado AI Act gets pared back
A working group is proposing to overhaul Colorado’s AI rules from broad mandates to a more targeted framework focused on transparency, notice, and human review—signaling a shift toward lighter, compliance‑friendly regimes. That retooling could influence state‑level expectations for municipal AI programs. (jdsupra.com)
Gov. Jared Polis announced the Colorado AI Policy Work Group reached unanimous support for a revised AI policy framework at a meeting on March 17, 2026. (governorsoffice.colorado.gov)) The work group’s draft explicitly removes mandatory bias-audit requirements, drops the law’s statutory “duty of care,” and would eliminate required impact assessments in favor of disclosure obligations, correction rights and mandatory human review for consequential automated decisions. (fisherphillips.com)) The original statute, Senate Bill 24-205, was enacted in 2024 and had been scheduled to take effect Feb. 1, 2026, before the legislature previously pushed the implementation date to June 30, 2026; the work group’s proposal would further delay a new effective date to Jan. 1, 2027. (skadden.com)) The work group was convened by Governor Polis and included representatives from consumer advocacy groups, hospitals, school districts and both large and small technology companies that met weekly to develop consensus language. (governorsoffice.colorado.gov)) Palantir Technologies cited Colorado’s AI regulations as a notable risk in its SEC annual reports and publicly moved its headquarters from Denver to Miami in February 2026, a relocation that state analysts say could cost Colorado hundreds of jobs and millions in economic output. (investors.palantir.com)) Legislators are now expected to use the work group’s consensus framework as the basis for a repeal‑and‑replace bill to be shared publicly in March 2026, a step stakeholders say aims to resolve months of standoff between industry and consumer advocates before the next legislative session. (bhfs.com))
