Your Medical Records Will Outlive Their Encryption: The Harvest-Now Threat to Health Data

A leaked password is an annoyance. A leaked diagnosis is forever. Medical records, psychiatric notes, and genomes are supposed to stay secret for decades, and that creates an awkward mismatch with the math protecting them. The keys guarding today’s hospital databases were chosen for a world without quantum computers, and they will not get a second chance.

Before you read: The NIST standards and the store-now-decrypt-later logic are covered in The Quantum Time Bomb: Why NIST’s New Encryption Standards Can’t Wait. This piece is about the sector where that logic hurts most, and why it is not a software update.

Table of contents

• The Shelf Life Problem
• Ransomware Doesn’t Need Quantum. It Needs Patience.
• The Implant Problem
• Where I Push Back
• Did You Know
• Sources

The Shelf Life Problem

Nobody designs a medical record to expire. Many US states require providers to keep patient records for at least seven years, and for children the clock typically runs until they reach adulthood plus several years more. A genome never expires at all. You can rotate a leaked password on a Tuesday afternoon. You cannot rotate the sequence you were born with.

That asymmetry makes ‘harvest now, decrypt later’ far worse in healthcare than in banking. A bank transaction loses most of its value once it clears. A medical history keeps its value for a lifetime, and sometimes for the lifetime of your children, whose DNA overlaps with yours. Encrypting a medical record today is like sealing a time capsule with a lock whose manufacturer has already announced it will stop working.

Ransomware Doesn’t Need Quantum. It Needs Patience.

Ransomware gangs stopped being satisfied with locking files years ago. Double extortion means they encrypt your systems and steal a copy first, then threaten to publish it. In February 2024, the ALPHV (BlackCat) group hit Change Healthcare, a claims processor that touches a huge share of US insurance payments. UnitedHealth later said the breach exposed data on roughly 190 million people. In May 2024, its CEO Andrew Witty told a House committee the company had paid about $22 million in ransom.

Here is the detail that should bother anyone who thinks quantum is the first problem. According to Witty’s testimony, the attackers got in through a remote-access portal that did not require multifactor authentication. The front door was a password, and no post-quantum cipher would have closed it. Anyone who has been logged out of a hospital EHR four minutes into a shift, then retyped a password off a sticky note, already knows how the post-quantum migration is going to go.

The quantum angle comes in afterward. A stolen archive is a fixed object. If the files inside were protected by RSA or elliptic-curve key exchange, the gang can keep the copy indefinitely and wait for a machine that can crack those keys. The ransom was the short-term payout. The archive is the long bet.

The Implant Problem

Pacemakers and implantable defibrillators are the most unforgiving security targets I know of. In August 2017, Abbott recalled about 465,000 pacemakers over a cybersecurity vulnerability that could let an attacker modify their programming and drain their batteries. The fix was a firmware update that had to be delivered at a clinic.

An implant placed today may run on a battery that lasts five to ten years before replacement, and the key that verifies its firmware signatures is baked in at manufacture. So a cipher that starts aging out in a decade has to protect a device that must keep working for a decade. Replacing the trust anchor of something sitting behind your ribs means a second procedure, and nobody signs up for surgery to update a root certificate.

Did you know that Dick Cheney’s doctors disabled the wireless feature of his implanted defibrillator in 2007 over fears of a targeted attack? He described it in his 2011 memoir, In My Time. It remains one of the more unexpected security reviews in medical history.

Where I Push Back

The quantum timeline is not the nearest danger to your medical file. The 2023 breach at 23andMe, which exposed profiles of roughly 6.9 million users, came from credential stuffing against accounts that lacked stronger protection. Stolen logins are still the cheapest way into health data, and they will remain so long after the algorithms change.

Still, the estimate for breaking RSA keeps shrinking. In 2019, researchers thought factoring a 2048-bit key would need around 20 million noisy qubits. By 2025, Craig Gidney had cut that number dramatically. That is why the timeline now sounds less like science fiction and more like a planning problem.

The US Department of Health and Human Services proposed in January 2025 to change encryption of electronic health information, at rest and in transit, from an ‘addressable’ implementation specification to a required standard under the HIPAA Security Rule. The proposal says nothing about post-quantum algorithms, and it is still a proposal.

My take: healthcare should start the migration now, because medical records outlive algorithms. But a hospital that does not enforce multifactor authentication has no business talking about quantum defenses yet. Lock the front door, segment the network, keep offline backups, and then start swapping the cryptography.

Did You Know

• In 2019, Gidney and Ekerå estimated that breaking RSA-2048 would take about 20 million noisy qubits and roughly eight hours. Gidney’s 2025 follow-up brought the qubit estimate below one million, with the runtime under a week.
• Pacemaker firmware updates are a rare case where a software patch can require a clinic visit, which makes every security fix a healthcare logistics problem.

Sources

• Andrew Witty testimony, House Energy and Commerce Committee (May 1, 2024)
• UnitedHealth Group disclosures on the Change Healthcare cyberattack (2024)
• HHS Office for Civil Rights, HIPAA Security Rule notice of proposed rulemaking, Federal Register (January 2025)
• FDA safety communication on Abbott (St. Jude Medical) pacemakers (August 2017)
• 23andMe security incident notice (October 2023)
• Craig Gidney and Martin Ekerå, ‘How to factor 2048 bit RSA integers in 8 hours using 20 million noisy qubits’ (2019)
• Craig Gidney, ‘How to factor 2048 bit RSA integers with less than a million noisy qubits’ (arXiv, May 2025)
• Dick Cheney, In My Time (2011)