Corrections and Updates
AI regulation changes quickly, and some of our earlier pages were not checked against the current text of the law as carefully as they should have been. When a claim about what a law requires, when it applies, or who it covers is wrong or out of date, we fix the page and record it here: what the page said, what it says now, and the primary source we checked it against.
Spotted something we missed? Write to contact@governome.com. Our editorial standards describe how we source and review what we publish.
Affected: EU AI Act Article 5, EU AI Act penalties
What it said: Said Article 5 prohibits eight AI practices.
What it says now: The consolidated text of Article 5 now lists ten: the original eight plus two added later, on AI-generated non-consensual intimate imagery and AI-generated child sexual abuse material. The consolidated text dates the second from December 2, 2026. Both pages now say so; we could not confirm the first one's application date or the amending instrument from the Official Journal.
Checked against: AI Act Explorer — Article 5 (consolidated text)
Affected: EU AI Act extraterritorial scope
What it said: Cited Article 2(4) as the exclusion for research, testing and development before an AI system is placed on the market.
What it says now: That exclusion is Article 2(8). Article 2(4) covers third-country public authorities and international organisations. The citation was corrected.
Checked against: AI Act Explorer — Article 2
Affected: Utah AI Policy Act
What it said: Said the rule that it is not a defense that generative AI made the statement is "now codified at Utah Code § 13-2-12".
What it says now: Section 13-2-12 was repealed on May 7, 2025, when SB 226 moved the AI consumer-protection provisions into a new chapter of Title 13. The page now says so and does not name a replacement section we could not confirm.
Checked against: Utah Legislature — Utah Code 13-2-12 (repeal note) and SB 226 (2025)
Affected: Why the US has no federal AI law
What it said: Said the House bipartisan AI task force's December 2024 report made 85 recommendations.
What it says now: The report contains 66 key findings and 89 recommendations. The figure was corrected.
Checked against: Bipartisan House Task Force on Artificial Intelligence — report (December 2024), as summarized by Baker McKenzie
Affected: The state preemption debate
What it said: Said the Commerce Department published its evaluation of "onerous" state AI laws on March 11, 2026, naming Colorado's statute. Also said the federal court paused enforcement two days after the Justice Department intervened.
What it says now: The evaluation was due March 11, 2026, but reporting through June 2026 says it had not been published, and Colorado's AI Act is named in the executive order itself. The pause came three days after the April 24 intervention.
Checked against: Federal Register — Executive Order 14365 (December 16, 2025)
Affected: FCC rules on AI robocalls
What it said: Said the political consultant behind the AI-cloned Biden robocalls faced 26 state criminal charges and described the FCC fine only as proposed.
What it says now: The FCC finalized the $6 million fine in September 2024, and press reports from June 2025 say a New Hampshire jury acquitted him on the state charges. We removed the charge count, which sources reported differently.
Checked against: Nextgov — Mastermind behind Biden AI robocalls fined $6M by FCC
Affected: Massachusetts AI regulation
What it said: Described the pending frontier AI provisions as applying only above a $500 million revenue and 10^26 FLOPs threshold.
What it says now: Those are the thresholds in New York's RAISE Act. We could not confirm them against the Massachusetts bill, which is in conference committee, so the page now says so.
Checked against: Massachusetts Legislature — Senate economic development bill (S.3178), in conference
Affected: Colorado AI Act, Colorado's AI Act timeline post
What it said: Described Colorado's SB 24-205 as in force from June 30, 2026, with a duty of reasonable care and annual impact assessments. The July post said only the effective date had moved.
What it says now: SB 24-205 never took effect. SB 26-189, signed May 14, 2026, repealed and reenacted it as an automated decision-making law, with developer and deployer duties beginning January 1, 2027. Both pages were rewritten, and we flag what we have not verified against the enacted text.
Checked against: Colorado General Assembly — SB26-189
Affected: EU AI Act Articles 9 to 15, conformity assessment and CE marking
What it said: Nine EU AI Act pages said high-risk obligations had been in force since August 2, 2026.
What it says now: The Digital Omnibus on AI (Regulation (EU) 2026/1744) moved the application date for standalone Annex III high-risk systems to December 2, 2027, and for high-risk systems embedded in Annex I products to August 2, 2028. Status labels, effective dates and an on-page timing note were corrected.
Checked against: AI Act Explorer — Article 26, application dates
Affected: Australia AI regulation
What it said: Described Australia's proposed mandatory guardrails for high-risk AI as a pending consultation likely to become legislation.
What it says now: The National AI Plan (December 2, 2025) dropped the standalone AI act and paused the mandatory guardrails. The page was rewritten around what applies now, including the Privacy Act's automated decision-making transparency rule from December 10, 2026.
Checked against: ABC News — National AI Plan (December 2, 2025)
Affected: NIST AI RMF, NIST AI RMF Govern function
What it said: Stated that Colorado's AI Act ties an affirmative defense to the NIST AI RMF.
What it says now: That defense belonged to the original SB 24-205. We have not confirmed that it carried over to SB 26-189, so the pages now say so instead of asserting it.
Checked against: Colorado General Assembly — SB26-189
Affected: United Kingdom, Canada, California
What it said: Pages lacked current status: no mention of the UK's promised Frontier AI Bill, no California effective dates, and Canada's page compared AIDA to a Colorado statute that had since been replaced.
What it says now: Current status and dates were added from the CPPA regulations page, UK Government material and law-firm commentary, and each page notes where it relies on secondary sources.
Checked against: California Privacy Protection Agency — Regulations
Affected: Texas TRAIGA, Utah AI Policy Act, Connecticut, Washington State, Massachusetts
What it said: Compared each state's approach to Colorado's "SB 205" duty-of-care model in the present tense.
What it says now: References now say Colorado's original act was repealed and replaced, and treat the duty-of-care model as the original approach rather than current Colorado law.
Checked against: Colorado General Assembly — SB26-189