OpenEvidence and Doximity clash in billion-dollar race to dominate AI for doctors
Summary
Two major players in the “ChatGPT for doctors” race — Sequoia-backed OpenEvidence and public company Doximity — have escalated a bitter legal fight into duelling lawsuits. OpenEvidence originally sued Doximity in June, alleging prompt-injection attacks and theft of proprietary code and trade secrets. Doximity has now counter-sued, accusing OpenEvidence of false advertising, harassment and employee poaching, and has moved to dismiss OpenEvidence’s claims.
The disputes touch on prompt-injection vulnerabilities, alleged misuse of physician identifiers, public exposure of patient records, and aggressive recruitment tactics. The outcome could set important legal precedents about trade secrets and competitive behaviour in the AI era for healthcare.
Key Points
- OpenEvidence sued Doximity in June, alleging prompt-injection attacks and theft of proprietary AI code and trade secrets.
- Doximity countered in September, accusing OpenEvidence of misleading advertising, harassment and poaching staff, and sought dismissal of the original suit.
- The cases centre on newer AI-specific issues such as prompt injection and whether exposing model system prompts amounts to trade-secret theft.
- Both companies are aggressively building physician-facing AI: OpenEvidence raised a $210m Series B at a $3.5bn valuation; Doximity acquired Pathway and is expanding Doximity GPT and free AI scribe tools.
- Legal rulings here could establish how courts treat alleged theft or misuse of model prompts, access methods, and other AI-era trade-secret claims.
Context and relevance
This is not just another tech lawsuit — it pits a fast-growing startup against a public healthcare platform in a space where accuracy, privacy and clinician trust are vital. As healthcare AI becomes central to clinician workflows, legal clarity on prompt injection, data access and trade secrets will shape how companies compete and how regulators and customers evaluate safety and compliance.
Author’s take
Punchy: This fight matters. If courts side with claims that prompt-extraction or impersonation can constitute trade-secret theft, it could chill certain competitive practices and force firms to rethink model access controls. Conversely, a dismissal could embolden aggressive data-collection tactics. Either way, expect ripple effects across healthcare AI startups and incumbents.
Why should I read this?
Because this story explains where healthcare AI and the law collide — and that could directly affect patient privacy, clinician tools and who wins the market for doctor-facing AI. We skimmed the court filings and the claims so you don’t have to; read this if you want to know what might change industry playbooks and legal standards next.