Colorado Law Mandating Therapists’ Real-Time Intervention During Client-AI Psychotherapy Sets Dubious Precedent

Colorado has become the first state in the nation to require that a licensed therapist be ready to intervene in real time whenever a client uses an AI-powered psychotherapy platform — a mandate that critics say could cripple the emerging field of digital mental health. On July 16, 2026, Colorado Governor Jared Polis signed Senate Bill 26-123 into law, imposing an unprecedented regulatory requirement on AI mental health providers. The law compels any company offering AI-led psychotherapy to have a qualified therapist monitoring sessions and able to step in at any moment if the AI struggles or the client shows signs of distress. The measure, touted as a patient-safety safeguard, has drawn sharp opposition from tech firms, psychologists, and industry analysts who warn it sets a dangerous precedent for regulating artificial intelligence in healthcare.

The AI psychotherapy sector has exploded since the pandemic, with millions of Americans turning to chatbots and virtual therapists for affordable, 24/7 mental health support. Startups like Woebot Health and Replika have attracted hundreds of millions in venture capital, while traditional platforms such as Talkspace and BetterHelp have integrated AI features. Yet the rapid adoption has outpaced regulation, and Colorado's new AI psychotherapy regulation is the most aggressive attempt yet to impose guardrails. The law applies to any entity that provides mental health treatment via AI to Colorado residents, regardless of where the company is based.

Under the law, practitioners must be licensed in Colorado and have direct access to the session feed, with the ability to interrupt the AI and take over the conversation immediately. The therapist must also review a sample of sessions post hoc and report any adverse events to the state medical board. Fines for noncompliance reach $10,000 per violation, and repeat offenders could face license revocation. The Colorado Psychological Association expressed mixed reactions: some members see the law as necessary to prevent harm, while others call it an unfunded mandate that will drive up costs and limit access for underserved populations.

The controversy centers on the feasibility of real-time therapist intervention. AI therapy platforms operate asynchronously or with minimal latency, and thousands of sessions can occur simultaneously. Requiring a live therapist for every session would essentially replicate the staffing model of traditional teletherapy, erasing the cost and scalability advantages of AI. Critics also note that therapists have limited ability to assess non-verbal cues through text or voice alone, and that the law could lead to liability confusion if an AI makes a mistake despite the therapist's oversight.

Industry observers say the Colorado AI psychotherapy law could have a chilling effect on innovation. Smaller startups may find the compliance costs prohibitive, and venture capital could shift to states with lighter regulation. Furthermore, the law may accelerate a patchwork of state-level rules, making it harder for national companies to operate. The American Psychiatric Association has called for federal guidelines instead, warning that Colorado's approach could set a 'dubious precedent' that other states follow.

Looking ahead, legal challenges are almost certain. Tech trade groups argue the law violates interstate commerce and oversteps the state's authority. Meanwhile, lawmakers in California and New York are watching closely, with similar bills already in draft. The outcome in Colorado will likely shape the future of AI regulation in healthcare for years to come. For now, therapists and AI companies are scrambling to understand their obligations — and whether this real-time intervention mandate is a safety net or a straitjacket for the next generation of mental health care.