Post quantum cryptography meets the medical device: a headline that overreads its data

Forescout's number, and what it actually measures#

A fox learns to distrust the loudest track in the snow, because the loudest track is rarely the freshest, and it follows the scent back to the source. The loud one this week is a headline that says critical medical devices cannot support the post quantum cryptography transition. Infosecurity Magazine ran it on 6 October, and the post below shows how quickly the wording drifted on the way to a social feed.

Read the original and the claim narrows. Forescout's research says only 6% of IoMT devices and 16% of OT devices currently use SSH implementations that support post-quantum cryptography, against 50% of IT devices. That is a measurement of one protocol on one date, not a verdict that these devices cannot be upgraded. The distinction matters because a security vendor who repeats the headline will be corrected by the first informed buyer.

Then the samples. The device figures come from a dataset of more than 2.5 million devices across more than 50 networks of healthcare delivery organisations, which are Forescout's own customers. A separate query of the Shodan search engine found over 5,500 exposed instances of 50 different medical information systems, and that is where the next figures come from. Of those exposed systems, only 31% support TLS 1.3, and 15% still support TLS 1.0 or 1.1. Two samples, two methods, one headline: always ask which dataset a figure belongs to.

• 6%

  of IoMT devices run post-quantum-capable SSH today

• 16%

  of OT devices in the same dataset

• 31%

  of exposed healthcare systems support TLS 1.3

• 2.5m

  devices in the underlying Device Cloud dataset

Why harvest now decrypt later suits a device that lives for a decade#

The threat that makes this a buyer's problem rather than a physicist's is harvest now decrypt later: an adversary records encrypted traffic today and waits until a quantum computer can read it. Forescout is careful here, and so should you be. Its report says the greatest near-term concern is not that quantum attacks are already happening. The concern is that data with a long shelf life, such as a patient record, will still be sensitive when decryption becomes possible.

Healthcare is the sector where shelf life is longest, and the devices are the slowest to change. A peer-reviewed comment in npj Digital Medicine reports that once certified, devices are rarely updated to meet new cryptographic standards. The imaging association COCIR adds a hint of how long the installed base lingers, noting that medical technology more than ten years old is outdated and hard to maintain. Equipment bought now can be in a ward in 2035.

None of that excuses delay, because the migration is a long trail through the undergrowth even when the threat is distant. The one hard cost anchor we found is the US federal estimate: the White House report put the work for prioritised federal systems at approximately $7.1 billion in 2024 dollars. It covers government systems, not hospitals, so treat it as a scale marker, not a forecast for healthcare.

The post quantum cryptography deadlines a device maker can date#

Some dates are real and some are rumours, so here are the ones with a document behind them. NIST published its first post-quantum standards in August 2024, and says of the key-encapsulation standard that ML-KEM is believed to be secure even against an adversary with a quantum computer, with a matching statement in the signature standard. NIST's transition draft, still a draft as of today, sets the line for older algorithms: 112 bits of security strength deprecated after 2030, disallowed after 2035.

Outside the US, the UK's National Cyber Security Centre asks organisations to carry out a full discovery exercise by 2028 and to complete migration of all systems, services and products by 2035. The EU's coordinated roadmap, from the NIS Cooperation Group, says that by 2035 the transition should be completed for as many systems as practically feasible.

Regulators for devices have not yet put a post-quantum date on the box. The FDA's final cybersecurity guidance, dated February 2026, gives recommendations on device design, labelling and premarket documentation, and section 524B of the Food, Drug and Cosmetic Act requires a plan to monitor, identify and address postmarket cybersecurity vulnerabilities and a software bill of materials. In Europe, MDCG 2019-16 asks that manufacturers demonstrate state of the art, and we found no post-quantum wording in it. A 2025 gap analysis found that both guidance documents have unclear scope and thematic gaps.

In the US, even the privacy rule is moving slowly: the proposed HIPAA Security Rule notes that experts anticipate that quantum computing will adversely affect digital communications, yet the HIPAA Journal reports a final rule postponed until July 2027.

Medical device security claims that stay true#

Now the marketing question, which is where folkfox earns its keep. Every security vendor selling into hospitals will start saying quantum-safe in the next year, and most of them will be saying more than the product can prove. Forescout names the reason itself: low-agility environments include regulated medical devices where changing a cipher can mean recertification or new hardware. A claim of readiness that ignores that is an invitation to be quoted back.

The honest rungs sell well, and they sit snugly in the hedgerow of what buyers already ask: what do we have, and what can we do about the part we cannot change? Forescout's own advice is to plan for PQC gateways and secure remote access in front of legacy systems, which is exactly the sort of concrete, checkable offer that a marketing page can state without flinching.

For a vendor, the content strategy follows from the claims ladder. Write the explainer on harvest now decrypt later for the chief information security officer, the inventory checklist for the biomedical engineer and the procurement language for the buyer, and keep the three voices distinct. Our cybersecurity marketing work and healthcare marketing work meet at exactly this point: regulated audience, evidence-led claims, long sales cycle. Related patterns appear in our piece on AI medical devices and the claim a regulator can quote back and in the ENISA threat landscape analysis.

None of this requires a shield, a padlock or a drumbeat of fear, and none of it needs a quarry of scare statistics. It requires the patience of a fox at a shuttered window: listen first, learn what is actually in the house, and speak only once you can say it plainly. The buyers in this field have heard every alarm, and they trust the vendor who sounds like an engineer.

Frequently asked questions#