DoD and DIU Announce Polygraph+: Modernizing Decades-Old Credibility Assessment Technology

In one of the most consequential developments in federal psychophysiological testing in half a century, the U.S. Department of Defense (DoD) and the Defense Innovation Unit (DIU) have announced a major initiative titled Polygraph+. The project represents a coordinated attempt to transition credibility assessment from legacy contact-based instrumentation into an AI-augmented, contactless standoff sensing paradigm.

Coordinated through the National Center for Credibility Assessment (NCCA) — a directorate of the Defense Counterintelligence and Security Agency (DCSA) — alongside the Office of the Undersecretary of Defense for Intelligence and Security (OUSD (I&S)), Polygraph+ brings together Silicon Valley commercial engineers, clinical biosignal manufacturers, and premier academic research laboratories. For investigative psychologists and polygraph practitioners worldwide, the announcement signals both an exciting technological evolution and a critical opportunity to examine how new sensing modalities intersect with validated forensic psychophysiology.

The Catalyst: Why Modernize the Polygraph?

The Department of Defense conducts thousands of credibility assessments each year. These examinations mitigate insider threat risk, inform high-stakes clearance decisions for access to Sensitive Compartmented Information (SCI), vet foreign sources, and support complex military criminal investigations. Yet, the foundational hardware of the conventional polygraph examination has remained largely unchanged for generations:

• Mechanical contact tethering: Convoluted rubber pneumograph chest assemblies to record respiration patterns;
• Electrodermal electrodes: Metal finger plates or adhesive Ag/AgCl sensors to record galvanic skin resistance/conductance;
• Cardiovascular occlusion: A continuous sphygmomanometer blood pressure cuff inflated on the upper arm, paired with a photoelectric plethysmograph on the thumb;
• Somatic activity monitors: Motion-sensor seat pads and arm pads to detect deliberate physical countermeasures.

While computerised data acquisition systems replaced mechanical ink pens in the early 1990s, the physical experience of being attached to pneumatic tubes, cuffs, and finger sensors remains intrusive. The blood pressure cuff, in particular, causes localized ischaemia and physical discomfort after 90 to 120 seconds of sustained pressure, limiting chart run durations and requiring frequent resting intervals.

Furthermore, the physical encumbrance can introduce artificial baseline anxiety — colloquially known as the “white-coat effect” — which unseasoned examiners may mistake for deception-related arousal if rigorous pre-test procedures and comparison-question controls are neglected.

“Advancements in standoff physiology sensing using cameras and thermal imaging have the potential to be less intrusive and more reliable than existing methods... Polygraph+ strives to deliver the transformative CA technology platform of the future.”
— Defense Innovation Unit (DIU) Official Statement

Core Technologies: Standoff Physiology and Optical Biosensing

Polygraph+ is designed to decouple physiological data collection from physical contact. Following an open commercial solicitation attracting more than 25 competitive proposals, the DIU and DCSA awarded prototyping contracts to two commercial pioneers:

1. Presage Technologies — Computer Vision and Remote Photoplethysmography (rPPG)

Presage has developed software platforms capable of extracting medical-grade physiological telemetry using standard consumer-grade digital cameras. Using advanced optical absorption algorithms — remote photoplethysmography (rPPG) — the system measures microscopic, imperceptible color fluctuations across facial micro-capillary beds as arterial blood pulses through the skin with each heartbeat. From these high-speed optical feeds, algorithms can infer instantaneous pulse rate, pulse wave velocity, respiratory sinus arrhythmia (RSA), and heart rate variability (HRV) without a single wire touching the examinee.

2. Altec — Precision Electrophysiology and Clinical Biosignal Analysis

Altec brings decades of experience in clinical neurophysiology, neuromuscular research, and sensor development. Their expertise centres on high-fidelity signal filtering, artifact removal, and the detection of subtle biosignals that reflect central and autonomic nervous system dynamics. In Polygraph+, Altec’s contribution bridges laboratory biomedical engineering with ruggedized operational environments.

3. Thermal Hyperspectral Imaging

In addition to optical video, the project harnesses infrared thermography to track instantaneous changes in surface blood perfusion. Under autonomic sympathetic activation, vasoconstriction occurs in peripheral facial zones (such as the tip of the nose and lips), while transient increases in blood flow occur in the periorbital and supraorbital vasculature — areas directly innervated by the sympathetic branch of the autonomic nervous system.

Academic Rigour and the DoD Ethical AI Framework

A frequent vulnerability of commercial “lie detector” startups has been the tendency to rush untested algorithmic products to market without peer-reviewed scientific validation. The DIU has addressed this risk by embedding an academic coalition of world-class research institutions into the Polygraph+ project:

• Massachusetts Institute of Technology (MIT): Pioneering computer vision, optical imaging, and sensor fusion architectures;
• Carnegie Mellon University (CMU): Global leaders in artificial intelligence, machine learning, and algorithmic fairness;
• University of Maryland: Foremost specialists in national security vetting, human language technology, and credibility assessment;
• Columbia University: Leaders in clinical biomedical engineering, cardiovascular physiology, and healthcare diagnostics.

This consortium is tasked with addressing three foundational hurdles:

1. Healthcare Industry Validation Standards

Contactless biosensors must demonstrate accuracy and test-retest reliability comparable to clinical gold standards (electrocardiography, continuous arterial blood pressure monitoring, and spirometry). If an optical camera estimates pulse or respiration, its error margins must be clinically quantified under diverse operational conditions.

2. Compliance with DoD Ethical Principles for AI

Adopted by the Department of Defense in 2020, these principles mandate that any artificial intelligence system deployed in defense missions must be:

• Responsible: Human operators exercise appropriate levels of judgment and care;
• Equitable: Algorithms are systematically audited to identify and eliminate bias across diverse examinee demographics (including skin tone, age, gender, and facial morphology);
• Traceable: The technical architecture provides explainable audit trails of how physiological signals were converted into feature scores;
• Reliable: Robust safety parameters prevent false positives caused by environmental lighting changes or unintentional physical twitches;
• Governable: Explicit human override mechanisms prevent automated life-altering clearance revocations without review.

3. Mitigation of Human and Evaluator Bias

Traditional manual chart analysis — while highly reliable when conducted by certified examiners following standardised numerical scoring protocols — still carries an element of subjective inter-rater variability. Polygraph+ aims to eliminate visual scoring ambiguity by automating feature extraction while preserving human-in-the-loop oversight.

The Investigative Psychology Perspective: Sensors Change, Mechanisms Remain

As an investigative psychologist and polygraph examiner, it is vital to separate sensor mechanics from psychological theory. Whether physiological activity is recorded through an inflatable rubber bladder around the biceps or an 8K thermal camera tracking periorbital perfusion, the physiological channel is merely the downstream consequence of a cognitive and affective event.

The polygraph does not measure “lies.” There is no unique, universal autonomic signature that occurs exclusively when an individual utters an untruth. As explained by Dr John J. Palmatier and Louis Rovner in Preliminary Process Theory, physiological differentiation across test questions arises from:

• Stimulus salience: How personally significant or threatening a question is to the examinee;
• Attentional capture & orienting response: The involuntary sensory and cognitive focus directed toward relevant versus comparison stimuli;
• Memory retrieval and recognition: The activation of episodic memory structures when confronted with specific knowledge;
• Autonomic mediation: Dynamic sympathetic activation and parasympathetic withdrawal in response to perceived jeopardy.

A superior optical camera records physiological changes with greater elegance and less physical burden. But it does not alter the psychological reality: credibility assessment succeeds or fails based on question formulation, examinee suitability, and structured procedural design.

If an ambiguous, loaded, or improperly phrased question is delivered to an examinee, the most sophisticated camera array in the world will merely record an exquisite, high-resolution artifact. Modern hardware cannot compensate for poor investigative methodology.

Screening vs Diagnostic Testing: Where Will Polygraph+ Excel?

In applied psychophysiology, the distinction between screening examinations and specific-issue diagnostic examinations is paramount:

1. High-Volume Security Clearance Screening

This is where Polygraph+ offers its most revolutionary operational impact. Federal agencies face massive backlogs in vetting personnel for initial clearance and continuous insider-threat monitoring. In a screening context, where dozens of broad questions regarding espionage, sabotage, or unauthorized disclosures are reviewed, contactless standoff sensing allows rapid, comfortable, high-throughput evaluations. Examinees do not waste time being strapped into harness hardware, and baseline readings stabilize far more rapidly.

2. Specific-Issue Forensic and Criminal Examinations

In single-issue criminal allegations — such as an asset theft, disputed statement, or serious disciplinary inquiry — the examinee faces specific jeopardy regarding a singular historical event. In this diagnostic domain, validated protocols such as the Comparison Question Test (CQT) or the Concealed Information Test (CIT) remain the gold standard. Until standoff systems undergo extensive, published field validation across thousands of real-world criminal case outcomes, conventional contact-sensor polygraph suites will continue to serve as the benchmark for evidentiary and legal proceedings.

The Indispensable Examiner: Why AI Cannot Replace the Clinical Mind

Peter Gust, Research Operations Specialist with DCSA’s Office of Research and Innovation, highlighted that the Polygraph+ project aims to push innovation into the wider Credibility Assessment and Interview Management community. The juxtaposition of “credibility assessment” and “interview management” is revealing.

A professional polygraph examination is an integrated diagnostic process, not an automated test. Over 80% of an examiner’s time and professional skill is expended before a single chart is recorded:

• Pre-test clinical suitability screening: Evaluating psychiatric history, neurodivergence, cognitive functioning, medication regimens, and cardiovascular stability to ensure the examinee is physically and psychologically fit to be tested;
• Psychological rapport and voluntary consent: Ensuring that the examinee understands the process, recognizes that participation is voluntary, and feels confident in procedural fairness;
• Exclusion of confusion and ambiguity: Dissecting the precise allegations and reviewing every single test question word-for-word so there is zero possibility of semantic misunderstanding;
• Exploration of collateral explanations: Understanding why an innocent examinee might harbour heightened sensitivity toward a particular topic, allowing the examiner to craft balanced, tailored comparison questions.

An algorithm cannot interview a traumatized witness, cannot reassure a falsely accused employee, and cannot evaluate whether a subtle nuance in question phrasing addresses the exact evidentiary dispute. The future envisioned by Polygraph+ is not examiner replacement; it is examiner empowerment through unobtrusive data collection and objective algorithmic support.

Implications for UK and Global Practice

Standards established by the NCCA and DCSA in the United States invariably influence forensic psychophysiology globally. In the United Kingdom, where polygraph testing is utilized under statutory frameworks for high-risk offender monitoring by the Ministry of Justice, as well as in corporate investigations and private disputes, the lessons learned from the Polygraph+ project will be followed closely.

At the Centre for Forensic Neuroscience, we view the DIU’s initiative as an affirmation of the scientific trajectory of our discipline. Credibility assessment is evolving rapidly from an isolated craft into a multi-disciplinary science uniting neuroscience, computer vision, autonomic physiology, and investigative psychology.

Conclusion

The Pentagon’s Polygraph+ initiative marks a welcome and necessary investment in modernizing credibility assessment technology. By replacing bulky, restrictive contact sensors with advanced standoff optical and thermal biosensors, the project has the potential to eliminate procedural discomfort, mitigate evaluator subjectivity, and vastly accelerate security vetting workflows.

Yet as the project unfolds over its two-year prototyping window at MIT, Carnegie Mellon, and Columbia, its ultimate success will depend on recognizing the boundary between sensor innovation and psychological science. Standoff sensors give us clearer windows into autonomic arousal, but it is the structured, ethical, and compassionate science of investigative psychology that allows us to understand what those physiological signals truly mean.

Selected References & Further Reading

1. Defense Innovation Unit (DIU) (2024). Department of Defense, Defense Innovation Unit Announce Polygraph+ Effort To Modernize Decades-Old Credibility Assessment Technology. Washington, D.C.: U.S. Department of Defense.
2. National Center for Credibility Assessment (NCCA) (2023). Federal Psychophysiological Detection of Deception Examiner Handbook. Fort Jackson, SC: DCSA.
3. Palmatier, J. J. & Rovner, L. I. (2015). Preliminary Process Theory: A comprehensive theory for understanding polygraph tests. International Journal of Psychophysiology, 95(1), 3-17.
4. Department of Defense (2020). Ethical Principles for Artificial Intelligence. Office of the Chief Information Officer, Washington, D.C.
5. American Polygraph Association (2021). Model Policy for the Evaluation of Polygraph Examination Validity and Reliability. APA Standards and Principles of Practice.

Dr Keith Ashcroft is a Chartered Investigative Psychologist and Certified Polygraph Examiner at the Centre for Forensic Neuroscience. The Centre provides evidence-based, confidential polygraph examinations and forensic behavioral assessments across the UK and internationally.