Sep 30, 2026

News

Your Caller ID Lied: What the Astrana Health Breach Teaches About Verifying Who's Actually Calling Your Practice

No malicious link. No infected attachment. No ransomware note on a locked screen. On September 22, 2026, Astrana Health — a healthcare management company whose network touches millions of patients through affiliated medical groups — told the SEC that attackers got into its systems by doing something almost embarrassingly low-tech: they called employees on the phone, pretending to be company personnel, with the caller ID spoofed to show the company's own main corporate number.

That's it. No exploit, no malware, no MFA bypass trick. Just a convincing voice and a phone number that looked right.

What actually happened

According to Astrana Health's own SEC filing, threat actors ran "a series of social engineering attempts in which threat actors, impersonating Company personnel and spoofing the Company's main corporate telephone number, contacted certain employees" to talk their way into system access. The company's security team caught the unusual activity, brought in outside forensic investigators, reset credentials, locked down access, and restored systems from clean backups. But by the time it was caught, "certain private and/or confidential information" had already been accessed or taken — potentially including patient data, employee records, credentialed provider information, and confidential business and financial records.

Astrana says it doesn't yet expect the incident to be financially material to the company. That's easy to say when you're a large, publicly traded healthcare organization with cyber insurance, a security team, and forensic investigators on retainer. It's a much harder sentence for a five-person practice to say about itself.

Why this is different from the last "just a phone call" story

We wrote a few weeks ago about the McKesson and Baxter breaches, where attackers called an IT help desk and talked their way into a password reset. That's one direction of the same problem: an outsider calling in, pretending to be someone on the inside.

Astrana Health's incident runs the other way. The attackers called out — to employees, pretending to be internal company staff — with caller ID spoofed to show a number the employees would recognize and trust. That's a meaningfully different failure point, because it doesn't target your help desk's verification process at all. It targets the instinct every employee has to trust a familiar name and number on their own phone screen.

Caller ID spoofing itself isn't new, but it's getting cheaper and more convincing. Voice phishing volume has risen sharply industry-wide over the past two years, and security researchers at CrowdStrike documented a 442% jump in voice phishing attacks in the second half of 2024 alone compared to the first half — a trend that has continued to climb since. The tools to fake a trusted number, and increasingly to fake a trusted voice, are now accessible to attackers who have no technical skill at all.

What this means if you're a South Florida practice

Most small practices assume vishing is something that happens to their IT vendor or their EHR company, not to their own front desk or billing staff. But the Astrana Health incident is a reminder that the target isn't always the practice's systems directly — it's confidential business, financial, and patient information sitting anywhere in the referral and vendor chain a practice touches. If your staff would grant access, reset a password, or read out a code over the phone to someone who sounded like "IT" or "the practice manager" and had the right number showing up, you have the same exposure Astrana Health did, just at a smaller scale.

This applies whether your practice is in Palm Beach, Broward, or Miami-Dade County, or among the practices we're increasingly working with as we extend our HIPAA-focused IT support into Martin and St. Lucie County. Caller ID spoofing doesn't care how big your practice is or which county you're in — it only cares whether your staff has a habit of verifying before acting.

The fix here isn't a new piece of security software. It's a policy, and a five-minute conversation with your staff.

Practical takeaways

  • Treat caller ID as informational, never as verification. It can be spoofed convincingly and cheaply — a name and number matching what's expected proves nothing on its own.

  • Put a callback rule in writing: if someone calls claiming to be IT, a vendor, or practice leadership and asks for a password reset, access change, or sensitive information, hang up and call that person or department back using a number you already have on file — never a number the caller gives you.

  • Never provide passwords, MFA codes, or remote-access approval over an inbound call you didn't initiate, no matter how legitimate the caller sounds or how much urgency they create.

  • Make sure your IT provider or internal team has a documented, staff-known process for identity verification on IT-related requests, and that every employee has actually seen it, not just the office manager.

  • Ask your IT partner whether your phone system supports STIR/SHAKEN caller ID authentication and spam/spoof flagging, and whether it's actually turned on.

  • Review who at your practice can approve access changes or password resets, and confirm that authority isn't something a convincing phone call alone can unlock.

Astrana Health had a security team, forensic investigators, and insurance, and the human layer still got through. For a small practice without those resources, the phone on your front desk is worth exactly as much attention as your firewall.

Get expert clarity

Frequently asked questions

Can my staff use the free version of ChatGPT or Claude at work?

Yes, for tasks with no patient information at all, like drafting a generic handout or a job posting. Never for anything touching a patient’s chart or PHI, per both companies’ own consumer terms.

What is a BAA and why does it matter for AI tools?

A Business Associate Agreement is a signed contract required under HIPAA before any vendor can touch PHI on your behalf. Anthropic and OpenAI only offer BAAs on business-tier or API plans, and it must be separately executed, not assumed from a paid subscription.

Does paying for ChatGPT Plus or Claude Pro make it HIPAA-compliant?

No. Both companies state that even paid individual consumer accounts (Plus, Pro, Max) are not covered by a BAA and PHI should not be entered into them.

What should our practice do before letting staff use AI with patient information?

Confirm a business-tier account is provisioned with an executed BAA, set a written policy distinguishing PHI from non-PHI use cases, and fold AI access into your existing audit-logging and access-review process.

How does UpsiteGroup help with this?

We confirm which AI tools your staff can access, verify a BAA is actually on file (not just technically available), and manage AI access controls alongside your EHR’s existing security configuration.

Build smarter systems and grow with confidence

Build smarter systems and grow with confidence