Five AI-Powered Attacks Every Payer Needs to Know

Health plans have a blind spot: AI-backed attacks working the voice and video channels members, providers, and staff already trust. Here are five ways attackers are using AI to threaten health plans:

1. Attackers use AI bots to attack the member services line

Fraudulent contact center conversations are not new, but the rapidly growing use of AI to automate call line reconnaissance and account takeovers fundamentally changes the stakes.

The chain is simple: a caller claims to be a patient or member, clears a knowledge-based authentication question built on data that’s already for sale, and either walks off with PHI or sets up an account takeover. Stolen identity packages sell for as little as $8 to $35 on dark web markets, and nearly 60% of organizations report fraudsters using that data to get past knowledge-based authentication (KBA). Fraudsters bypass KBA in more than half of attempts and clear OTP challenges roughly 25% of the time.

AI transformed that attack into a volume business. AI-driven attacks grew 1,390% over six quarters.¹ At the customer level, that’s an average of 287 AI attacks a day, a 6,400% jump from the year before.² Automated bot calls now account for more than half of all fraud at some major healthcare organizations. And many of these calls are not overtly malicious, at least not at first. A bot often conducts initial reconnaissance, mapping out the IVR and validating the data it has. Once it has identified a clear path and potentially collected additional information about the patient, it can prepare for a more direct attack, like an account takeover.

Listen to this real example of a bot call:

When these attacks work, the damage hits fast: PHI and sensitive data are released, payments are rerouted, funds are drained from accounts. Perhaps most costly of all is the loss of trust and reputation, especially in an industry where trust is increasingly difficult to preserve.

2. Fake job candidates apply for open roles

Health plans hire remote workers by the thousands: IT and engineering roles, care management staff, member services reps, etc. Remote hiring is a prime space where deepfake candidates thrive.

In 2025, the FBI released an official public service announcement warning about North Korean operatives trying to get hired at U.S.-based companies. Related reports affirm that healthcare is increasingly targeted by these attacks. These operatives and other bad actors use stolen identities and real-time face-swap tools to build a resume, a LinkedIn^® profile, and an interview performance convincing enough to clear a screen.

The exposure for health plans is direct. A fraudulent hire in claims, IT, or member services receives legitimate day-one access to PHI, member data, and claims and payment platforms. And the volume of exposure continues to rise. Pindrop’s own 2026 hiring pipeline analysis measured 1 in 47 applicants as having ties to North Korea, a 630% increase from the year before.

If you haven’t detected a North Korean or deepfake candidate, it’s probably because you haven’t been looking.

3. Executive impersonators authorize wire transfers

Payers collect and move large sums constantly: provider reimbursements, capitation payments, vendor payments, etc. That makes finance and executive teams a direct target for impersonation.

A finance employee at Arup joined a video call with who looked and sounded like the company’s CFO and wired $25 million before anyone realized the call was a deepfake. WPP’s CEO was impersonated over WhatsApp with a cloned voice and deepfake video in an attempt to pressure an executive into releasing funds. Attackers exploit the instinct to move fast and trust the person who outranks you, putting all healthcare organizations, including health plans, at risk.

We deepfaked our CFO

Watch this video to see how easily attackers can join meetings as a fake executive

The damage from an executive impersonator goes beyond manipulated employees. Fake videos of leaders making statements about financials or product updates can do serious damage to brand reputation, too. Leaders have access, influence, and trust. AI lets scammers mimic all three.

4. Fake providers or vendors submit fake claims and invoices

Payers work closely with a web of third parties: TPAs, PBMs, provider networks, claims clearinghouses. Every one of those relationships runs on emails, calls, and invoices that are increasingly easy to fake. Fraudulent activity in these relationships is not new, but as with the other attacks in this article, AI has elevated the risk at hand.

The U.S. Department of Justice released a report describing charges against 324 defendants connected with over $14.6 billion in alleged fraud. The report describes charges against several defendants who allegedly “used artificial intelligence to create fake recordings of Medicare beneficiaries purportedly consenting to receive certain products.” The beneficiaries’ confidential information was then used to fuel $703 million in fraudulent claims to Medicare, of which $428 million was paid.

One cybersecurity firm reported attackers using AI to build realistic email threads and fake invoices that looked legitimate enough to get approved and paid before anyone caught the vendor wasn’t real. This playbook is industry agnostic and, thanks to AI, can easily scale. Humans catch AI-manipulated content correctly only about half the time, no better than a coin flip.

Statistically, enterprises cannot rely dependably on staff ears or eyes alone to detect fraudulent activity. An attacker posing as a trusted supplier just needs to send a realistic email, a cloned voice message, or a fake invoice to break through defenses.

5. Bad actors target IT helpdesks to reset credentials and access critical systems

Using stolen information, including employee ID numbers, names, and department information, attackers social engineer IT helpdesks, with or without the help of AI, to access PHI, financial information, and more.

MGM Resorts is the case every security team already knows. Attackers found an employee’s LinkedIn profile, called the help desk pretending to be that employee, and talked their way into a password reset. One convincing phone call led to unauthorized access across MGM’s systems and a multi-day outage. Microsoft recently warned users about a similar tactic spreading through Teams, with attackers posing as IT staff to request remote access before moving through internal systems.

Securing employee credentials and access, while also offering quick support to legitimate staff, is difficult to balance. As the pool of stolen information grows, and as generative AI tools make it easier to look and sound trustworthy, the risk of data and financial exposure through helpdesk manipulation continues to expand.