Will AI Replace Call Center Agents? What the Labor Data and Research Show for Health Systems

What You Need to Know

• Current evidence points to AI handling routine patient calls rather than replacing call center agents. Staff keep insurance, billing, clinical and complex calls.
• AI is changing tasks far more often than it cuts jobs¹. About 95% of US firms using AI, across all industries, reported no employment change from AI over six months¹.
• Sending routine calls to AI lets a health system absorb call volume growth without proportional hiring, with staff time kept for calls that need a person.

Will AI replace call center agents in health care?

AI is more likely to handle routine patient calls than to replace call center agents, based on current evidence. AI can finish scheduling, rescheduling, cancellations, confirmations and common questions. Agents keep the calls that need judgment: insurance, billing, refills, complex scheduling, clinical questions and callers who want a person.

For a health system, the practical effect is capacity. Sending routine calls to AI lets call volume grow without proportional hiring.

Why can't a regional health system hire its way through call volume growth?

Three things make hiring alone hard to scale with call volume: replacement hiring, growing demand and after-hours calls.

Replacement hiring never stops. The Bureau of Labor Statistics (BLS) expects all receptionist and customer service openings to come from replacing workers who leave²,³. Recruiting continues even when a team is fully staffed.

Demand keeps growing. Private health care and social assistance is projected to add more jobs than any other sector through 2035⁴. When a system adds providers and visits, calls to book, change and confirm visits can grow with them.

Calls arrive after hours. The University of Arkansas for Medical Sciences (UAMS) call center closed at 5 p.m.⁵ Patients who called after hours to cancel left messages that waited until the next morning⁵.

Compare the cost of adding capacity by hiring, outsourcing or automating. Our guides to medical receptionist cost, medical receptionist turnover, healthcare call center cost and healthcare call center outsourcing cover those costs.

What does research show so far about whether AI will replace call center agents?

Studies so far find AI changing tasks more often than cutting jobs, although projections show fewer jobs by 2035. Nearly all US firms using AI, across all industries, reported no employment change from it¹. One warning sign is slower hiring of young workers in the most AI-exposed occupations, although AI hasn't been shown to cause that pattern⁶. None of these studies covers health care call centers.

Projections. BLS projects customer service representative jobs down 5% from 2025 to 2035, partly from automation². For receptionists, it projects a 2% drop³. Both projections are national and cover all industries, not health care alone²,³. Yet BLS warns that its methods aren't designed to capture "extremely rapid technological change"⁷.

Firms. The Census Bureau surveyed US firms in all industries from late 2025 to early 2026¹. Far more firms used AI on tasks than changed employment¹.

Firms using AI across more business functions were somewhat more likely to report employment declines¹. The survey can't show cause¹.

Earnings and hours. In Denmark through mid-2024, AI chatbots had no significant effect on earnings or recorded hours in 11 exposed occupations, including customer support⁸.

Hiring. US payroll data through June 2026 show a gap for workers ages 22 to 25 in the most AI-exposed occupations⁶. Their employment was 19% below where it would be had it kept pace with less-exposed peers⁶. Slower hiring drove the gap, not more people leaving or losing jobs⁶.

The data show a pattern, not a cause⁶. After accounting for education, the gap roughly halves and in one sample is no longer statistically significant⁶.

Inside a contact center. An AI assistant helped chat agents at one software company resolve 15% more issues per hour on average⁹. Newer and less experienced agents gained most, while the most skilled agents saw small gains in speed and small declines in quality⁹. Most of the agents worked outside the US⁹.

Which patient access calls can AI absorb, and which stay with staff?

AI can absorb routine, high-volume calls such as scheduling, rescheduling, cancellations, confirmations, common questions and after-hours requests. UAMS began using an AI system to let patients cancel appointments after hours⁵. It later expanded the system to all day⁵.

Staff keep insurance coverage questions, billing and payment questions, prescription refill requests, complex or non-standard scheduling, clinical questions and escalations. BLS expects some companies to keep in-house service centers "particularly for complex inquiries such as refunding accounts or confirming insurance coverage"².

A scheduler's call can still matter for high-risk patients. One safety-net internal medicine practice flagged some appointments as high risk for a no-show¹⁰. Flagged appointments that got a scheduler's call plus reminders had 33% no-shows, versus 36% with reminders alone¹⁰. The no-show count included same-day cancellations¹⁰.

Routing routine calls to AI lets a health system absorb volume growth without proportional hiring while staff time stays on complex calls. The split also needs escalation rules and a plan to bring staff into the change. Our AI receptionist vs. human receptionist guide compares the two staffing models call by call.

How do Commure AI Call Center Agents help a health system absorb call volume growth?

Commure AI Call Center Agents take the routine share of patient calls, letting a health system absorb call volume growth without proportional hiring. Commure reports typical results of a 30%+ reduction in inbound calls handled by your team. Your staff keep complex calls, escalations and patients who need a person.

The division of work is set during implementation. Commure Agents handle scheduling, rescheduling, cancellations, confirmations, intake and common questions, while routing rules set up for your organization send other calls to the right team by intent. Complex or non-standard requests go to staff. The agents capture messages as structured tasks for follow-up.

The agents answer calls 24/7/365 in English and Spanish, including after hours. During surges, they take many calls at once without queueing or busy signals.

Every call is recorded, transcribed and scored. To review how the work divides, use the dashboard, which tracks resolution rate, scheduling success, abandonment and patient sentiment.

To see which of your calls fit, request a complimentary call center health analysis. It breaks down your call volume by type, length, category and outcome, then models the automation opportunity.

How should a call center leader plan for AI on patient access lines?

Plan for AI on patient access lines by starting with your own calls. National data can't say whether AI will replace call center agents at your health system, but your call mix shows how much volume growth AI can absorb.

1. Sort a week of calls by reason and outcome. A complimentary call center health analysis can do it for you, using a week of call transcripts and basic call data under a signed BAA.
2. Mark which call reasons need a person. Start with insurance coverage, billing and payment, prescription refills, complex scheduling, clinical questions and calls in languages other than English and Spanish.
3. Project next year's call volume. Use your call data to plan how much growth AI can absorb without proportional hiring. Keep enough staff capacity for complex calls.
4. Define escalation paths before go-live. Decide which calls go to schedulers and which go to clinicians, including how urgent clinical calls reach the on-call provider. Write what callers hear during the handoff.
5. Decide how staff will review AI-handled calls. Check the same weekly healthcare call center metrics for AI-handled and staff-handled calls. Start with scheduling success and abandonment.
6. Brief staff and labor partners early. Explain which calls are moving to AI and how staff work will change. If your call center staff are unionized, federal or state labor law and your union contract may require notice and bargaining before their work changes¹¹. Set the timeline with labor counsel.
7. Check AI disclosure, call recording and outbound calling rules with counsel. The rules come from state AI laws, recording laws and federal calling rules.- California: The rule covers hospitals, clinics, physician offices and group practices that use generative AI for written or spoken messages about patient clinical information¹²,¹³. They must add a disclaimer, which is spoken at the start and end of a voice call¹²,¹³. They must also tell patients how to reach a human provider or other appropriate person¹²,¹³. Messages a licensed or certified provider reads and reviews are exempt¹²,¹³. Scheduling and billing aren't covered¹²,¹³.- Utah: A business must tell a caller who asks that they're talking with generative AI¹⁴. State-licensed professionals must also say so at the start of a call when the AI collects health data or gives medical advice¹⁴.- Texas: Health care providers must disclose AI used in relation to health care services¹⁵.- Call recording: Although federal law allows recording when one party to the call consents¹⁶, some states, including California, require every party's consent to record a confidential call¹⁷. Callers may need to hear a recording notice.- Outbound calls: The FCC treats AI-generated voices as "artificial" under the Telephone Consumer Protection Act (TCPA)¹⁸. Outbound AI calls must state who is calling at the start and give a callback number¹⁸,¹⁹. They may also need the patient's prior consent and an opt-out option, depending on the call's purpose and the number called¹⁸,¹⁹.