Age verification by conversational AI operators, $25,000 in statutory damages by digital imitation, mandatory clinical examination in care decisions.
A regulatory review published on 2 August 2026 identifies the wave of texts adopted by the US states in July, which shift the regulation of AI from the general framework to the general public.
Colorado replaced its automated decision-making system with SB 26-189, whose obligations will only come into effect on January 1, 2027. The HB 26-1263, or Chatbot Safety Act, on the other hand, requires conversational AI operators to estimate the age of their users, to report that it is an AI, to provide protection against explicit content for minors and to submit an annual report to the Attorney General. HB 26-1139 requires a clinician's review in AI-assisted management decisions and restricts the direct delivery of psychotherapy by an automated system.
Digital Imitation and Facial Recognition
In Hawaii, Act 247 creates civil liability for digital imitations generated by IA without authorization, with statutory damages of up to $25,000 per room, while Act 248 supervises company AIs: disclosure of the artificial nature of the interlocutor, crisis resources, protection of minors and prohibition of presenting himself as an approved mental health professional. Virginia, for its part, has tightened restrictions on facial recognition by local police. At the federal level, Decree EO 14409, which has been operational since 1 August, commits AI cybersecurity projects and a voluntary framework for early access to border models.
Read the full article on the original source
Go to the source →