RI Doctor Says AI Medical Firm Fired Him After Flagging FDA and HIPAA Risks
Friday, November 28, 2025
A new federal lawsuit in Rhode Island pits a Brown University Health pediatric psychiatrist, Dr. Bhargav Patel, against an AI healthcare company marketed as Sully.AI. Patel alleges he was terminated after raising concerns about potential FDA and HIPAA noncompliance. The company, registered in Delaware and headquartered in California, did not respond to a request for comment at the time of publication.
The core allegation
Patel claims that while serving in roles that culminated in his appointment as Chief Medical Officer at Odiggo Inc. d/b/a Sully.AI, he flagged compliance issues and was later fired. The suit states he worked months without pay, was misclassified as a contractor, and was terminated after objecting to what he believed were misleading regulatory claims on sales calls. The defendants' business centers on AI-powered medical software for physicians and hospitals in the U.S. and abroad.
What Patel says happened
- Recruitment: Executives pitched a "win-win" arrangement with flexible compensation. Patel began work around March 11, 2024, and was added to the company's Slack on March 29, 2024.
- Early contributions: Over March-November 2024, he says he worked 5-10 hours weekly on marketing, clinical templates, platform consultation, and subscriber growth, claiming more than 500 meetings booked and ~$15,000 in annual clinic revenue.
- Compensation shift: He alleges no pay until November 22, 2024, then $5,000/month as a contractor. On January 31, 2025, he became a W-2 employee at the same rate, plus benefits and ~$30,000 in options.
- Promotion: On July 22, 2025, Patel says he became Chief Medical Officer at $14,000/month with ~79,029 restricted shares (~$250,000 in equity). He also helped recruit a head of engineering.
The FDA claim dispute
The suit describes an August 4, 2025 sales call with a hospital regarding an "AI radiologist." Patel says he was told internally the product was essentially a ChatGPT wrapper. On the call, however, an executive allegedly claimed the model was FDA approved.
When the hospital's representative said they could not find evidence of an FDA-approved AI radiology model, the call reportedly ended abruptly. Patel says he warned leadership that representing a product as FDA approved without clearance could be illegal and misleading, and messaged co-founders and team members to be careful with compliance claims.
Termination and legal claims
Patel asserts those concerns triggered a breakdown in the relationship and his termination. The six-count suit seeks back pay and damages, including claims related to wages and retaliation. The complaint frames him as a whistleblower pointing to FDA and HIPAA risks.
Why this matters for healthcare leaders
AI vendors are moving fast, but clinical buyers carry the regulatory risk. Misstating FDA status or overlooking HIPAA safeguards can expose hospitals and practices to significant penalties and reputational damage.
- Demand specifics: Ask vendors for the exact FDA pathway (e.g., 510(k), De Novo) and the product's unique device identifier if applicable. Verify claims against public listings.
- Separate clearance from aspiration: "FDA registered" or "Class I exempt" is not the same as "FDA cleared" or "FDA approved." Pin down the exact claim in writing.
- Lock down HIPAA: Confirm a signed BAA, data flow diagrams, de-identification methods, retention policies, and subcontractor controls.
- Audit proof: Require a compliance packet-model documentation, intended use, validation data, known failure modes, and monitoring plans.
- Go-live gates: Tie deployment to proof of clearance (if required), security testing, risk assessment, and clinician oversight plans.
Quick reference
About Sully.AI (as described in the suit)
The complaint says the company develops and supports AI-powered medical software to increase physician efficiency and efficacy. Its marketing tagline: "Sully is the most integrated superhuman team of AI-employees for healthcare," citing adoption by organizations with over 100,000 providers. The suit names multiple executives as defendants.
Note: These are allegations in an ongoing lawsuit. Facts may be disputed, and no court has ruled on the merits.
If you're building internal skills for evaluating AI tools-from clinical safety to vendor due diligence-see AI courses by job role for structured training paths.
Your membership also unlocks:
