Clinical AI is no longer one category. Some tools answer questions. Some write notes. Some summarise records. Some route patients. The new competitive frontier is whether the AI sits close enough to the clinician's task to influence what happens next — not just what the clinician knows, but what the clinician does.
Doximity has just crossed that line into prescribing.
Doximity announced in its Q4 FY2026 earnings on 13 May 2026 that it reached over 800,000 active prescribers using its workflow tools — a 30% year-on-year increase and the largest engagement jump the company has ever reported. Nearly half of those providers used clinical AI tools last quarter, and prompts per user nearly doubled from January to April alone. The company announced partnerships with Aledade (bringing the Clinical AI Suite into thousands of value-based primary care organisations) and Photon Health (enabling in-workflow e-prescribing for the first time). Over 1,000 prescribers participated in the e-prescribing beta.
The CEO described the feature as allowing doctors to write prescriptions "in a few taps" after telehealth or while on the go, with the patient choosing their preferred pharmacy from their phone. Full-year FY2026 revenue was $644.9 million, up 13%. The company generated a record $107 million in quarterly free cash flow. 140 health systems are now buying the Clinical AI Suite.
This is not a chatbot launch. This is a platform thesis: clinical AI becomes most valuable when it collapses steps in the clinician's day — from question to note to prescription to pharmacy routing — within one professional workflow.
What Is Doximity Prescribe?
Doximity Prescribe is best understood as an in-workflow prescribing layer: a way for clinicians to issue prescriptions more quickly from a trusted professional platform, with patient-facing pharmacy choice and Photon-powered infrastructure in the background. The clinician selects the medication within the Doximity Clinical AI Suite. The patient receives notification and chooses their preferred pharmacy. The prescription is routed through Photon's infrastructure to the selected pharmacy.
Critical distinctions: this is not "AI decides the prescription." It is not autonomous prescribing. It is not the same as Utah's AI renewal pilots (where AI may eventually submit renewals without individual physician review). It is workflow compression — reducing the friction between clinical decision and prescription fulfilment within the platform clinicians already use for communication, documentation, drug search, and clinical AI queries.
Doximity's advantage is distribution. With 85% of US physicians registered on the platform and 800,000+ active prescribers, Doximity already owns the professional attention of the majority of US physicians. Adding prescribing to an ecosystem that already includes Scribe (documentation), Ask (clinical AI), PeerCheck (physician review), and Dialer (telehealth/communication) creates a workflow loop that competing standalone tools cannot replicate.
Why Photon Matters
Photon Health raised $16 million to build end-to-end prescription infrastructure — not merely e-prescribing transmission, but the consumerisation of the prescription handoff. Photon's platform includes prescribing and routing infrastructure, a pharmacy partner network, a consumer marketplace showing real-time price and stock information, delivery options, and capabilities including prior authorisation and clinical decision support.
Traditional e-prescribing sends the prescription to a pharmacy. Modern prescribing infrastructure helps determine where, how, and whether the patient can actually obtain the medication — including pharmacy choice based on price, stock availability, delivery options, and coverage. AI can then sit around drug selection assistance, coverage verification, prior authorisation automation, pharmacy routing optimisation, and adherence monitoring.
The Photon integration means Doximity is not just adding a prescribing button. It is adding a prescription fulfilment layer that extends from the clinical decision through to the patient obtaining the medication — with the patient making informed pharmacy choices based on real-time price and availability data.
Why This Is Bigger Than E-Prescribing
The clinician does not want another browser tab. The winning interface is likely to be the one that joins the clinical question, the consultation note, patient communication, prescribing, follow-up, documentation, and evidence trail into one workflow. Doximity is not launching a feature — it is demonstrating a product thesis about what happens when clinical AI platforms reach sufficient scale and workflow breadth.
The strategic sequence is revealing: Doximity started with physician identity and professional networking. Added communication (Dialer, messaging, fax). Added drug reference (drug search, monographs via Pathway acquisition). Added documentation (Scribe). Added clinical knowledge (Ask/DoxGPT, PeerCheck with 10,000+ physician reviewers). Added telehealth. Now added prescribing (Photon). Each layer makes the next more valuable because the clinician stays within one platform — and each additional workflow the platform supports increases the daily usage occasions that create retention.
The Safety and Governance Question
Prescribing remains a regulated clinical act — regardless of how frictionless the workflow becomes. Drug choice requires diagnosis, contraindications, allergy checking, pregnancy status, renal and hepatic function assessment, interaction checking against concurrent medications, monitoring requirements, and patient preference. Pharmacy routing and price transparency are operationally helpful, but they do not remove clinical accountability.
AI-adjacent prescribing tools should be evaluated differently from clinical knowledge tools. When the AI makes it easier to prescribe, the question becomes not just "did the AI answer correctly?" but "what clinical action did the AI make easier, and who remains accountable for it?" The workflow compression that makes prescribing faster also compresses the verification window — the time between deciding and executing — which is where safety checks happen.
The article should distinguish between different categories of "AI prescribing": AI-informed prescribing (the AI provides clinical information that informs the prescriber's decision), AI-enabled e-prescribing (the AI platform transmits and routes the prescription the clinician has chosen), AI prescription renewal (the AI generates renewal recommendations for existing prescriptions — as in Utah's pilots), and autonomous prescribing (the AI makes the prescribing decision independently — not yet authorised anywhere for new prescriptions).
Doximity Prescribe sits in the second category. The clinician decides. The platform executes. The distinction matters for governance, liability, and clinical safety.
Why UK, Canadian, and Australian Clinicians Should Care
The pattern is portable even if Doximity is US-focused. UK primary care faces similar prescribing friction: repeat prescribing consumes substantial administrative time, EPS routing has its own inefficiencies, Pharmacy First pathways are creating new prescribing workflows for pharmacists, shared-care prescribing requires coordination between primary and secondary care, and structured medication reviews generate deprescribing and switching decisions that need to be actioned efficiently.
Canadian family medicine faces appointment-access pressure and provincial prescribing workflows where administrative burden contributes to physician burnout and practice closure. Australian GP practice faces similar prescribing, pharmacy integration, and HealthPathways coordination challenges.
The lesson: clinical AI will increasingly move from answering questions to facilitating actions — and the governance, safety, and accountability questions must move with it.
Where iatroX Fits
iatroX sits upstream of the prescribing act: guideline-grounded answers, drug information via SmPC/eMC, clinical calculators, CPD documentation, and exam-grade learning. The clinical knowledge layer that informs the prescribing decision — not the prescribing transaction itself.
Doximity Prescribe suggests that clinical AI is leaving the search box and entering the transaction layer. For clinicians, the key question is no longer only "Did the AI answer correctly?" but "What clinical action did the AI make easier, and who remains accountable for it?"
Use iatroX for the clinical knowledge behind prescribing decisions →