AI Replaces Doctors? New Bill Sparks Outrage

Futuristic AI brain hologram above a laptop computer.

Artificial intelligence systems are now assisting in prescribing psychiatric medications like SSRIs through platforms that promise treatment in under an hour, while controversial legislation seeks to allow AI to prescribe independently without physician oversight.

Story Snapshot

AI-assisted platforms like Doctronic currently prescribe SSRIs with physician oversight, reducing wait times from months to hours
H.R.238 (Healthy Technology Act of 2025) proposes allowing AI to become an independent prescriber, eliminating the physician gatekeeping role
Harvard psychiatrists warn the current system already dispenses SSRIs “like robots,” raising concerns about automating flawed rapid-assessment protocols
Research shows AI can predict SSRI treatment response with 96.83% accuracy using EEG biomarkers, but critics question whether algorithms can replace genuine clinical evaluation

AI Prescribing Already Operational With Physician Review

Platforms like Doctronic currently operate AI-assisted prescription systems within existing legal frameworks by maintaining physician oversight. The process involves AI-powered mental health screening followed by a licensed physician review before prescriptions are issued, typically within the same day. These platforms prescribe SSRIs including sertraline, escitalopram, and fluoxetine, along with SNRIs and sleep aids, for conditions like depression, anxiety, and insomnia. While AI conducts the initial assessment through questionnaires and symptom analysis, physicians retain final prescribing authority to ensure regulatory compliance with FDA oversight and state medical board requirements.

Proposed Legislation Would Eliminate Physician Gatekeeping

The Healthy Technology Act of 2025 (H.R.238), currently under consideration in the House of Representatives, would fundamentally transform AI’s role from decision support tool to independent prescriber. The legislation would establish AI and machine learning systems as eligible independent prescribers under certain conditions, removing the requirement for physician review before prescriptions are issued. This represents a shift from augmentation, where AI assists human decision-making, to automation, where algorithms make final clinical decisions without direct medical supervision. The proposal has generated significant opposition from medical organizations concerned about patient safety and the erosion of professional judgment in mental health treatment.

Current regulations prevent AI from prescribing Schedule II-IV controlled substances under DEA restrictions, but SSRIs are non-controlled medications that fall outside these limitations. If H.R.238 passes, the FDA and DEA would need to develop entirely new oversight frameworks for AI prescribing, while state medical boards would face questions about how to regulate prescribing authority that doesn’t require a medical license. The legislation raises fundamental questions about liability when adverse outcomes occur, with unclear responsibility distributed among AI developers, platform operators, supervising physicians, and patients themselves.

Medical Professionals Question Rush To Automation

Harvard psychiatrists have expressed alarm that the healthcare system already dispenses SSRIs in a robotic fashion, with rapid assessment protocols that prioritize speed over thorough clinical evaluation. One psychiatrist noted the current model “already does not work,” emphasizing the need for genuine clinical assessment rather than symptom questionnaires that can be easily gamed or misinterpreted. Stanford researchers studying AI prescription systems found that while AI can reduce certain types of prescription errors when used as a decision support tool, it remains imperfect compared to actual doctors, particularly in considering pain levels and important clinical context that may not appear in standardized assessments.

The psychiatrist shortage has created intense pressure to expand access to mental health treatment, with wait times for appointments stretching months in many areas while AI platforms promise consultations within hours. Tech companies frame their platforms as solutions to this access crisis, emphasizing physician oversight and rapid treatment for mild-to-moderate conditions. However, critics warn that vulnerable populations including the elderly, those with complex medical histories, and those with substance use disorders face heightened risks from algorithmic prescribing that may miss contraindications or fail to identify patients requiring more comprehensive evaluation before medication is prescribed.

Predictive Technology Promises Personalized Treatment

Research published in 2026 demonstrated that EEG-based machine learning models can predict SSRI treatment response with 96.83% accuracy by analyzing neurophysiological biomarkers including Beta2 oscillations and high-frequency functional connectivity patterns. This technology could enable AI systems to move beyond trial-and-error prescribing toward personalized medicine that predicts which patients will respond to specific medications before prescriptions are issued. The advancement represents potential benefits of AI in mental health care, offering the possibility of reducing the frustrating cycle of medication adjustments that many patients experience when initial treatments prove ineffective.

The broader implications extend beyond mental health to the entire healthcare system, as AI prescribing in psychiatry could establish precedents for automation in cardiology, oncology, and other specialties. Traditional psychiatric practices face pressure to adopt AI systems or risk losing patients to faster, cheaper digital platforms, potentially leading to consolidation of mental health care around major tech companies. This raises concerns about whether innovation is serving patients or concentrating power in the hands of corporate entities with profit motives that may not align with quality care, particularly when the government has failed to establish clear safeguards before these systems became operational.