Doximity Shares Surge as Clinical AI Outperforms Anthropic in Medical Accuracy
A specialized drug reference system drove a massive stock rally, signaling a shift toward vertical AI in healthcare.
Doximity shares experienced a massive rally this week after the company revealed its specialized Clinical AI significantly outperformed a leading general-purpose model in medical accuracy. The surge underscores a growing investor belief that domain-specific AI is superior to general LLMs for high-stakes clinical environments.
According to CEO Jeff Tangney, Doximity's Clinical AI recorded a 4.8% error rate, starkly contrasting with the 13.6% error rate seen in Anthropic's Fable 5 model. This performance gap triggered a significant stock surge, with pre-market gains reported between 70% and 100%. The rally comes as a sharp reversal for the professional networking platform, which had recently seen its shares tumble following a miss in fourth-quarter earnings expectations.
The Vertical AI Advantage
Tangney attributed the accuracy advantage to two key differentiators, most notably a built-in, expert-verified drug reference system. This system is designed specifically to ensure dosing accuracy and perform rigorous interaction checks, tasks where general-purpose models often struggle due to a lack of verified, real-time medical grounding. By augmenting the AI with proprietary, expert-validated data, Doximity has positioned its tool as a reliable utility for physician workflows rather than a general chatbot.
To scale this technology, Doximity has integrated its Clinical AI Suite into thousands of independent medical practices. These deployments were made possible through strategic partnerships with Aledade and Photon, moving the technology from a theoretical benchmark to active clinical use.
Implications for Healthcare Tech
This shift in market sentiment suggests that "vertical AI"—models trained or augmented with specialized, proprietary data—may be the only viable path for AI adoption in medicine. While frontier models from companies like Anthropic offer broad capabilities, the high cost of medical errors makes a 13.6% error rate unacceptable in a clinical setting. Doximity's ability to slash that rate to under 5% demonstrates that specialized grounding can decouple a company's valuation from short-term financial misses, as investors pivot their focus toward long-term technical superiority.
What to Watch
As Doximity continues to roll out its AI suite through its partner network, the industry will be watching to see if these accuracy gains translate into measurable reductions in physician burnout and improved patient outcomes. While the stock has reacted positively to the benchmarks, the long-term challenge remains whether this technical edge can consistently offset the financial headwinds seen in the company's recent earnings reports.