The Impact Of AI Voice In Patient-Doctor Interaction

Hui Sang Yun is Co-founder and CEO at Endo Health. Medical doctor turned engineer, backed by General Catalyst and a16z speedrun.

The patient-doctor relationship is broken. Over 30% of clinic time is spent on administrative tasks, and AI is already cutting that by a third. Yet, technology could be the game-changer. While text-based communication has failed to address the root issues of this relationship, AI-powered voice could be the interface that finally fixes it.

Recent advancements in AI voice technology have shifted the frontier. OpenAI’s Realtime Voice APIs and ElevenLab’s models can now clone human voices with near-perfect intonation using just 15 seconds of audio. These AI voices process real-time audio loops with near-human latency and decreasing cost per minute. While smart speakers primed consumers, the challenge now is creating a magical experience to deliver healthcare with this new “why now” moment of technology.

Why “Voice” Is the Next AI Interface

Voice has become a dominant feature in consumer tech, and now it’s ready for healthcare. AI-driven voice has moved beyond simple speech recognition to mimic human-like interactions. Today, you can interact with an AI voice assistant that sounds almost indistinguishable from a real person. This leap in technology isn’t just about conversation, it’s about improving the way we connect emotionally, understand intent and manage complexity in patient care. The question isn’t if voice AI will enter healthcare, but when.

Current Pain Points In Patient-Doctor Communication

Healthcare has long struggled with communication barriers. The reliance on electronic health records (EHRs) and asynchronous portals drains valuable time from patient-doctor interactions. Patients often feel unheard, leading to frustration. Chronic care programs using text-based portals often see response rates below 50%, and real-world adoption of digital health tools remains low, with many solutions struggling to surpass 10%–20% active usage. Meanwhile, rural and aging populations still rely heavily on phone calls, and access gaps persist.

Why Voice Outperforms Text For Care

Voice builds emotional connection and trust that text rarely achieves. Behavioral studies show patients are significantly more forthcoming over voice than text, often disclosing more personal details and engaging more deeply in care conversations. This triggers trust and rapport in a way that text can never achieve. Voice captures nuances like background noise, breathing and emotional tone, offering a richer layer of clinical data. These subtle cues can make a difference in care delivery, especially in triage and diagnosis.

Early Movers

The potential of voice AI in healthcare is no longer theoretical. Clinics are increasingly automating patient inbound calls with HIPAA-compliant AI voice engines. New consumer-facing products can create a phone-based AI weight loss coach to deliver personalized nutrition guides. These products deliver a scalable, 24/7 phone calling experience in healthcare, something that was previously impossible.

The Case For Chronic Care

Chronic conditions such as diabetes, hypertension and weight loss require consistent, daily touchpoints. AI voice can seamlessly integrate into these workflows by offering automated nudges that don’t disrupt the patient’s routine. These conditions present low-acuity, high-frequency needs, perfect for AI intervention. With its ability to triage, escalate and document in one continuous interaction, voice has the potential to revolutionize how we monitor and manage chronic care, and democratize the level of care that only the top 0.1% had access to before.

Roadblocks

Despite the promise, there are several obstacles. Adoption is a big one: patients aren’t yet comfortable talking to bots, and past negative experiences with Siri or Alexa make them hesitant to try. Building trust takes time, and voice interfaces need to evolve from ambient documentation to direct interaction. Privacy concerns are also critical: AI voice interactions must comply with HIPAA regulations, and we must navigate consent for call recordings and data residency. Reimbursement is another hurdle, as CMS only pays for care if specific CPT codes or chronic-care management bundles exist. Lastly, who’s liable if an AI-driven suggestion goes wrong? These are the challenges that must be tackled to ensure widespread adoption.

What’s Next—And What You Can Do

The next two years will be pivotal. As adoption data accumulates, expect regulatory clarity and updates to CPT codes, unlocking broader reimbursement pathways. The startups that win will be those that master the real-time voice stack and billing systems. Health systems, payers and founders should act now, as voice AI is crossing the credibility gap faster than EHRs did a decade ago. The phone in every pocket can become the most scalable care channel we’ve ever had, but only if we build for trust, transparency and real outcomes.

Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?