# When Your Body Becomes the Attack Surface
A pacemaker is a networked computer implanted in a human chest. An insulin pump is a networked computer attached to a human abdomen. A cochlear implant is a networked computer wired into a human auditory nerve. A brain-computer interface is a networked computer threaded into the human brain itself.
Each of these devices processes data, communicates wirelessly, receives firmware updates, and connects to external systems for monitoring and management. Each has an attack surface. Each can be compromised.
This is not a future scenario. Pacemakers, insulin pumps, and cochlear implants are deployed in millions of patients today. Brain-computer interfaces are in human clinical trials. The cybersecurity implications are present-tense for the first category and near-future for the second. Neither is speculative.
The Planetary Defense Model covers all of it. Not because CDA designed the PDM to address medical devices and neural interfaces, but because the PDM describes the architecture of defense, and that architecture does not change when the "system" being defended is a human body instead of a corporate network. The six domains apply. The question is how.
The FDA estimates that over 190,000 medical devices are on the U.S. market, and a significant percentage of devices manufactured since 2015 have wireless connectivity. Insulin pumps communicate via Bluetooth. Pacemakers and implantable cardioverter-defibrillators (ICDs) communicate via proprietary wireless protocols to bedside monitors that relay data to the manufacturer's cloud platform. Continuous glucose monitors transmit readings to smartphones. Infusion pumps connect to hospital networks for dosing orders and telemetry.
Each wireless interface is an attack surface. Each network connection is a data flow that carries sensitive health information. Each firmware update mechanism is a potential vector for malicious code delivery.
These are not theoretical vulnerabilities. Researchers have demonstrated practical attacks against implanted medical devices:
Pacemaker/ICD attacks. Security researchers (including the late Barnaby Jack, who demonstrated the attack before his death in 2013) have shown that implantable cardiac devices can be wirelessly accessed from distances of up to 50 feet, that patient data can be extracted, that device settings can be modified (including disabling the device or delivering unauthorized shocks), and that the wireless communication protocols use little or no encryption or authentication. The FDA issued safety communications regarding cybersecurity vulnerabilities in specific cardiac device models from multiple manufacturers.
Insulin pump attacks. Researchers have demonstrated the ability to wirelessly command insulin pumps to deliver unauthorized doses, potentially causing life-threatening hypoglycemia. Johnson & Johnson issued a safety warning to patients using its Animas OneTouch Ping insulin pump after researchers demonstrated that the device's wireless communication could be intercepted and commands could be replayed.
Hospital network attacks. Ransomware attacks on hospitals have disrupted medical device operations by compromising the hospital networks that connected devices depend on. During the 2017 WannaCry attack, hospitals in the UK's National Health Service were forced to divert ambulances and cancel surgeries because medical systems, including imaging equipment, were encrypted. The 2024 Change Healthcare attack disrupted healthcare operations nationwide, affecting payment processing, prescription fulfillment, and insurance verification.
Medical device security is uniquely difficult because of four constraints that do not apply to conventional IT systems:
Patient safety. A security patch that causes a medical device to malfunction can kill the patient. The risk of leaving a vulnerability unpatched must be weighed against the risk of a patch causing device failure. This is not a theoretical trade-off. Medical devices undergo rigorous FDA certification, and firmware changes require recertification in many cases. The patching cadence for medical devices is measured in months to years, not days to weeks.
Device lifespan. A pacemaker implanted today will remain in the patient's body for 7 to 15 years. The software running on that device will still be running in 2040. The wireless protocols it uses will still be the same. The encryption algorithms it employs may be deprecated before the device reaches end-of-life. The device cannot be easily "upgraded" because it is surgically implanted.
Resource constraints. Implanted devices operate on batteries that must last years. Cryptographic operations consume power. Complex authentication protocols consume power. The security controls that are standard on a laptop (AES-256 encryption, TLS 1.3, certificate-based authentication) may be computationally or power-prohibitive on a device the size of a matchbox running on a battery for a decade.
Regulatory complexity. Medical device cybersecurity sits at the intersection of FDA regulation (device safety), HIPAA regulation (data privacy), and cybersecurity best practices that were designed for IT systems, not implanted medical devices. The FDA's premarket and postmarket cybersecurity guidance has strengthened significantly since 2023, but the regulatory framework is still catching up to the threat landscape.
Brain-computer interfaces (BCIs) are devices that create a direct communication pathway between the brain and an external system. Neuralink (Elon Musk's company) received FDA approval for human trials in 2023 and implanted its first human patient (Noland Arbaugh) in January 2024. Synchron's Stentrode device, which is inserted through a blood vessel rather than requiring open brain surgery, has been in human trials since 2022. BrainGate, a research consortium, has demonstrated BCI-controlled cursor movement and robotic arm operation in paralyzed patients.
The current use case is medical: restoring communication and motor control for patients with paralysis, ALS, and other neurological conditions. The long-term trajectory includes cognitive enhancement, direct brain-to-brain communication, and brain-to-computer data transfer that bypasses keyboards, touchscreens, and voice entirely.
When a computer is wired into a human brain, every cybersecurity concern becomes existential.
Neural data sovereignty. A BCI reads brain signals: electrical patterns that encode thoughts, intentions, motor commands, and potentially emotional states. This is the most personal data conceivable. Who owns it? Where is it stored? Who can access it? Can a BCI manufacturer use neural data for research? Can law enforcement compel access to neural data with a warrant? Can an insurer use neural data to assess risk? These are DPS questions (Data Protection and Sovereignty) with stakes that no previous data type has approached.
The SDP (Sovereign Data Protocol) principle, "Your data lives where you decide. Period," takes on a literal meaning when the data is a recording of a person's brain activity. Neural data sovereignty is not a compliance checkbox. It is a human rights question.
Interface attack surface. A BCI communicates wirelessly with external devices (a phone, a base station, a cloud platform). That wireless interface is an attack surface. If the interface can be compromised, an attacker could potentially read neural data (surveillance of thoughts), inject signals (manipulating perception, motor control, or emotional state), or disable the device (denial of a capability the patient depends on for basic communication or motor function).
The VSD question: what can an attacker see, and how do we shrink what they can reach? For a BCI, the answer has physical and neurological consequences that no conventional cybersecurity vulnerability carries.
Device integrity. A BCI runs firmware that interprets neural signals and translates them into commands. If that firmware is compromised (through a malicious update, a supply chain attack on the device manufacturer, or direct exploitation of the wireless interface), the device's interpretation of neural signals could be altered. The user thinks "move cursor right." The compromised device interprets "move cursor right" as "type a message the attacker dictates."
This is an SPH concern (firmware integrity, configuration management, baseline behavioral monitoring) with consequences unique to neural interfaces: the compromised "system" is a human brain.
Neural authentication. BCIs may enable a new form of authentication: neural patterns as biometric identity. The way a person's brain processes a specific stimulus (viewing a familiar image, recalling a specific memory) produces a signature that is unique, difficult to forge, and theoretically stronger than fingerprints or facial recognition.
Neural authentication is an IAT advancement with a significant caveat: unlike a password, a neural pattern cannot be rotated if compromised. Unlike a hardware key, a neural interface cannot be replaced (without surgery). The permanence of neural biometrics creates a risk profile unlike any existing authentication factor.
The Planetary Defense Model's six domains apply to the human body as a cybersecurity theater without modification. This is the structural proof that the PDM is architecture, not a technology catalog.
DPS (Data Protection and Sovereignty). Neural data, biometric data, continuous glucose readings, cardiac telemetry, infusion pump dosing history: this is the most sensitive data any system processes. Classification: Restricted, at minimum. Encryption: mandatory at rest and in transit. Sovereignty: the patient must control where their biological data lives and who can access it. SDP applies in its most literal form.
VSD (Vulnerability and Surface Defense). Every wireless interface on a medical device is attack surface. Every Bluetooth connection, every proprietary RF protocol, every cloud API endpoint that receives device telemetry. CSR applies: every surface you expose is a surface we eliminate. For implanted devices, minimizing the wireless attack surface (reducing broadcast power, limiting connection windows, requiring proximity authentication) is the primary VSD control.
SPH (Security Posture and Hygiene). Firmware integrity, device configuration, behavioral baselines. A pacemaker that begins transmitting data at unusual intervals or a BCI that begins processing signals differently from its baseline may indicate compromise. APC applies: the device's posture must be monitored continuously, and deviations from baseline must trigger investigation. The challenge: firmware updates for implanted devices carry patient safety risk, so the patching cadence that SPH normally enforces must be balanced against clinical risk.
IAT (Identity Access and Trust). Who can access the device? The patient. The cardiologist. The device manufacturer's remote monitoring platform. Each accessor needs authenticated, authorized, time-limited access. The manufacturer's access should not persist between service appointments. The cardiologist's access should not extend to the manufacturer's analytics platform. ZPA applies: trust nothing, possess nothing, verify everything. The device manufacturer should be able to service the device without possessing the patient's neural or cardiac data.
TID (Threat Intelligence and Defense). Detection of attacks against medical devices and BCIs. Anomalous wireless communication patterns, unexpected firmware behavior, unauthorized connection attempts. PDI applies: threat intelligence specific to medical device attack techniques (which are documented in ICS-CERT advisories and FDA safety communications) must inform the detection rules. The watchtowers must watch the body.
RGA (Risk Governance and Assurance). FDA cybersecurity requirements (premarket and postmarket), HIPAA privacy and security rules, medical device certification standards (IEC 62443 for connected devices), neuroethics governance frameworks for BCIs, liability frameworks for device manufacturers whose products are compromised, and insurance models for cyber-physical harm. PCA applies: compliance with medical device cybersecurity requirements is not an annual audit. It is a continuous state enforced by the device manufacturer, the healthcare provider, and the patient.
The convergence of cybersecurity and human biology is not a distant future. It is a present reality for medical devices and a near-future reality for neural interfaces. The organizations and individuals responsible for defending these systems, device manufacturers, healthcare providers, patients, regulators, and cybersecurity practitioners, need frameworks that address the full scope of the challenge.
The PDM provides that framework. Not because it was designed for medical devices. Because it was designed to describe how defense works at every scale, for every system, in every theater. The body is a theater. The six domains apply. The methodologies apply. The operational discipline applies.
What changes is the stakes. A compromised laptop loses data. A compromised pacemaker loses a life. The architecture of defense is the same. The consequences of failure are not.
Word count: 2,298