Verification with Quantum Sensing

Over the last couple of years, quantum research has moved beyond basic research in several areas, with quantum technologies now covering a wide range of devices and applications, from secure communication and computing to sensors using atoms, defects in diamond, or light. Some of these sensors are promoted for detecting objects over long distances or in noisy environments, including personnel, vehicles and facilities in the field. While they were not designed with arms control verification in mind, some may eventually support verification or lower its costs. Sensors generate observations that must be interpreted and combined with other evidence before states will accept them. A device that is more stable, selective, or deployable can shift a measurement from the laboratory to the field, from destructive sampling to noninvasive scanning, or from a fixed instrument to a mobile robot or drone. The key policy question is whether the sensor provides useful, validated information under realistic verification constraints. For the technologies discussed here, the evidence is promising in some areas but does not yet establish verification readiness.

In this text we discuss the pros and cons of the applications that appear most promising for future verification uses. First, we briefly describe what makes quantum sensing different from traditional sensing. We then highlight potential use cases before examining hurdles that may affect willingness to adopt new verification technology. We conclude that quantum technology is unlikely to revolutionize verification and that any practical benefit must be demonstrated against established alternatives.

Verification Frame: What Counts as Useful Evidence?

Verification is more complex than detection alone. Detection asks whether an instrument sees a signal; verification asks whether that signal can support a judgment about compliance with a treaty, declaration, inspection mandate, or agreed confidence-building measure. Hence, a useful instrument must be linked to an inspection concept: what is being verified, where the measurement is allowed, what reference data are available, what level of uncertainty is acceptable, who handles the data, and what happens if the result is disputed. Even a technically strong result is not usable if the parties have not agreed to the instrument, procedure, and evidentiary role.

This framing favors incremental use cases. Quantum gravimetry and magnetometry can reduce a search area or corroborate declared layouts. Quantum-assisted inertial navigation can document inspection routes and measurement locations where satellite navigation is unavailable. Nonlinear-interferometer spectroscopy can screen materials, gases, or residues for follow-up analysis. Each application still requires a legal basis, clear chain of custody, independent calibration, host-state protections, and confirmatory methods. The state of the art is best described as pre-verification: field or platform trials exist in adjacent applications, but the sensor classes discussed here have not been documented in treaty inspections.

How Quantum Sensors Help Solve Existing Verification Problems

Quantum sensors exploit controlled quantum states to measure external fields or optical properties with high precision. Cold atoms can act as matter-wave interferometers for gravity, acceleration, and rotation; atomic vapors and NV centers in diamond can serve as magnetometers; and parametric down-conversion in nonlinear crystals can link illumination at one wavelength with detection at another. These approaches are attractive because many verification-relevant signatures are weak: a tunnel slightly changes the local gravity gradient, a buried metal casing perturbs a magnetic field, and a chemical compound absorbs only at characteristic wavelengths. These technologies are dual-use: the same advances can serve civil, defense, or verification applications.¹

The main contribution of this technology is to improve parts of the measurement chain. Quantum sensors may lower uncertainty, reduce drift, make a field measurement more repeatable, or provide a more selective screening signal. They must still deal with terrain, vibration, temperature drift, atmospheric absorption, sampling bias, adversarial denial and concealment, as well as operator error. Their policy value lies in broadening the toolbox available for verification tasks, but they cannot determine which measurement questions are worth pursuing in the field, which data can be shared or filtered, or which results require confirmatory laboratory analysis. Any claim of improvement must therefore be tested at the level of individual tasks and instruments, not inferred from the quantum label.

Quantum-Based Gravimetry: Supporting Local Underground Surveys

Quantum gravimeters and gravity gradiometers measure tiny variations in gravitational acceleration or in how that acceleration changes over short distances. Because gravity cannot be shielded, dense structures and hollow spaces can create measurable anomalies. In a controlled field test above a known utility tunnel, a cold-atom gravity gradiometer reported 1.5 to four times better statistical performance than commercial survey gravimeters, although overall repeatability was similar once systematic effects were included.² It was a static test on a known target, and each survey point took approximately ten minutes.

For verification, possible applications include the local search for undeclared tunnels, cavities, shafts, large buried objects, or geophysical changes associated with underground nuclear testing. Detection depends on target size, depth, density contrast, sensor stability, survey geometry, and corrections for terrain and geology. Satellite gravimetry is valuable for large-scale Earth-system monitoring, but its spatial resolution is far too coarse to identify individual facilities or submarines.³ The most realistic benefit is improved site-survey evidence, particularly when gravimetry is complemented by ground-penetrating radar, seismic methods, maps, and visual inspection.

In verification terms, gravimetry is most valuable as an anomaly-finding and corroboration tool. It can help inspectors decide where to focus more intrusive methods, while interpretation still requires contextual evidence. Any deployment would need pre-agreed survey grids, terrain models, uncertainty thresholds, and rules for handling geophysical data that may reveal unrelated infrastructure. Limited inspection time makes survey speed as important as sensitivity.

Quantum-Based Magnetometry: Local Magnetic-Anomaly Mapping

Magnetometers measure disturbances in magnetic fields. Quantum magnetometers based on atomic vapors, superconducting devices, or nitrogen-vacancy centers in diamond can be extremely sensitive. In security debates, they are often associated with submarine detection, but magnetic signatures decrease rapidly with distance and can be masked by geology, infrastructure, vehicles, cables, and natural field variations.

The near-term verification value is local and inspection oriented. Magnetometry can help locate buried metallic components, cables, casings, drums, or equipment; check whether declared layouts match observed magnetic anomalies; or support on-site inspection methods that already include magnetic field mapping or other geophysical techniques in certain phases and according to defined procedures.⁴ The potential quantum advantages are improved sensitivity, stability, vector measurement, and miniaturization. However, the literature reviewed here does not provide a verification-specific field comparison against high-end classical magnetometers. In many tasks, classical instruments may in fact be sufficient.

As evidentiary method, magnetometry is most effective when used to guide sampling, excavation, visual inspection, or other geophysical measurements. A verification protocol should therefore specify what counts as a reportable anomaly, what reference measurements are needed, and how inspectors avoid overinterpreting clutter in industrial or military environments.

Quantum-Assisted Inertial Navigation: Documenting Inspections When External Signals Fail

Inertial navigation determines the position of a vehicle or inspection platform by measuring its acceleration and rotation to pinpoint the location relative to an initial position. It does not need GPS, but small errors accumulate over time. Cold-atom accelerometers, gyroscopes, and clocks may reduce that drift and improve navigation where satellite signals are unavailable, jammed, spoofed, or not permitted. Public UK flight trials in 2024 and Royal Navy trials in 2025 show that relevant components can operate on mobile platforms, but the published reports do not provide an end-to-end performance comparison with conventional inertial navigation.⁵

For arms control verification, the most relevant use is inspection autonomy and documentation. Using inertial navigation, robots, vehicles, drones, or inspectors could navigate inside buildings, tunnels, ships, or cluttered sites where GPS is unavailable, compare actual routes and spaces with declared design information, and document where measurements were taken. This could improve repeatability, reduce reliance on host-provided infrastructure, and make subsequent reviews of inspection results more robust.

The verification caveat is that navigation data may itself constitute sensitive information. A high-resolution route record may reveal facility layouts, access constraints, or process flows. Quantum-assisted navigation would require map integration, classical inertial systems, procedural safeguards, host-state review where appropriate, and human oversight. It should be treated as part of the evidence-management system, not merely as a convenience for inspectors.

Nonlinear Interferometers: Chemical and Biological Screening with Light

Nonlinear interferometers are a particularly promising but still emerging quantum sensing option for verification. In a conventional interferometer, a beam splitter creates two paths that are later recombined. In a nonlinear interferometer, special crystals create linked light paths through parametric down-conversion. One path can pass through the sample at a useful wavelength, while the other can be measured with a detector that is easier to operate. Because the two paths are linked, the measured light carries information about what happened in the sample path even when the sample wavelength itself is not detected directly. This principle emerged from induced-coherence experiments and was later extended to imaging of light that is not directly detected.⁶

The arms control relevance of this technology comes from wavelength separation. Many chemicals and biological materials have strong and selective absorption features in the mid-infrared fingerprint region. Conventional mid-infrared spectroscopy is powerful, but its sources and detectors are often costly, noisy, bulky, or require cooling. Nonlinear interferometry can probe samples in the mid-infrared while using mature visible- or near-infrared detectors. Reviews and experiments have demonstrated the feasibility of infrared spectroscopy with visible readout, gas spectroscopy using near-infrared detection, near-infrared grating-spectrometer readout of mid-infrared spectra, biological and hyperspectral microscopy with this approach, and ultra-broadband quantum infrared spectroscopy.⁷ In 2024, an open-path experiment detected methane below 4 parts per million and common solvent vapors at roughly 9–14 parts per million in ambient air.⁸ These figures show technical progress, but not yet an advantage over portable Fourier-transform infrared (FTIR) spectroscopy or readiness for inspections.

For chemical weapons verification, the most realistic application is a screening and triage instrument rather than a replacement for accredited laboratory analysis. During routine or challenge inspections, nonlinear-interferometer spectroscopy could help screen declared chemicals, precursors, solvents, degradation products, residues, exhaust streams, or air samples before resorting to more intrusive sampling. This should be framed carefully: OPCW-related analytical practice and training already includes FTIR, portable FTIR, Raman spectroscopy, and GC/MS for rapid identification and verification support.⁹ Handheld infrared capability alone is therefore nothing new. The possible added value lies in combining mid-infrared probing with visible or near-infrared readout, lower detector burden, tailored spectral windows, and greater compatibility with managed-access screening.

For biological verification and investigations, the potential is more limited and should be described carefully. The Biological Weapons Convention (BWC) does not have a standing verification regime comparable to the Chemical Weapons Convention (CWC). Consequently, nonlinear-interferometer microscopy or spectroscopy would be relevant mainly to investigations, confidence-building, biosafety inquiries, or future verification concepts rather than routine BWC inspections.¹⁰ Mid-infrared microscopy can distinguish broad classes of biomolecular material, such as lipids, proteins, carbohydrates, or culture media, and can reveal spatially resolved chemical patterns. Its value would lie in serving as a nondestructive screening and prioritization method: flagging samples, surfaces, or production residues for follow-up analysis, or supporting biosafety investigations in contexts where low light exposure and label-free measurement matter.

Potential Verification Uses and Maturity

These developments justify targeted research and carefully designed field trials. High-gain and loss-compensated schemes may improve sensitivity or robustness in selected configurations, yet most systems remain laboratory instruments rather than rugged field devices.¹¹

Verification Design: From Instrument Readings to Accepted Evidence

A verification system is a political construct and thus more than a sensor. It includes the inspection mandate, managed-access rules, data handling, calibration, chain of custody, and dispute-resolution procedures. Such a regime typically verifies declarations of treaty-controlled substances, such as chemical weapons compounds or nuclear fissile material, and checks for any undeclared activities that might contravene an agreement. New technological methods therefore also have a political dimension that must be addressed before use. Technical validation alone does not guarantee acceptance: inspected states may reject a novel instrument because of mandate, confidentiality, safety, precedent, or other political concerns. Quantum sensing should therefore be introduced through joint validation involving verification parties and vendors, rather than through isolated technology demonstrations. Across all domains, trials should test not only sensitivity but also false positives and negatives, operating time, data security, operator training, ruggedness, and inspector usability.

Illustrative pilots would make policy recommendations more concrete. For gravimetry and magnetometry, mock sites should include known voids, pipes, metallic clutter, and realistic terrain corrections. For inertial navigation, trials should test whether quantum-assisted navigation improves route reproducibility without exposing sensitive facility details. For nonlinear interferometers, tests should use safe chemical simulants, mixtures, humidity changes, and realistic background conditions, with field readings subsequently compared with established infrared spectroscopy and mass-spectrometry-based laboratory methods.

Managed access is important for all quantum sensors because a measurement can reveal information—such as facility layouts, buried infrastructure, route data, or industrial process signatures—that extends beyond the inspection question. This is particularly important for spectroscopic sensors. A host state or facility operator may object that a broad-spectrum instrument could reveal proprietary or militarily sensitive information unrelated to the inspection mandate. Nonlinear interferometers offer a design advantage here: the probe wavelength, detected wavelength, spectral range, and reference library can be engineered. In principle, a device could be built as a broad spectrometer for laboratory investigations or as a restricted-range instrument that only tests treaty-relevant bands. Physical filters, locked software, cryptographic logging, jointly held calibration samples, or jointly operated classifiers could reduce unnecessary data collection while preserving evidentiary value. The same filter logic can also be applied to classical photonic techniques.

The distinction between technology and data must be explicit. For decision-making, the crucial issue is not the full disclosure of proprietary sensor designs, but whether the data are reliable, properly calibrated, traceable, and generated in accordance with certified procedures. A verification organization does not need to reveal every technical detail of an instrument if performance has been validated by agreed test protocols, calibration standards, tamper-evident logging, and independent review. This is a general verification principle rather than a quantum-specific exception.

Access to technology is also likely to be mediated by commercial suppliers. A sharp technical divide between states is not the central concern at this stage; many systems will be provided by specialized firms, spin-offs, or start-ups rather than built domestically by each verification actor. Commercial quantum gravimeters and companies developing cold-atom sensors, quantum-diamond magnetic sensors, and compact cold-atom hardware already exist.¹² The policy task is therefore to certify instruments, publish nonsensitive performance data, enable shared training, and avoid dependence on opaque vendor claims.

The most important policy step is to define evidentiary thresholds before deployment. Inspectors, host states, and verification organizations should know whether a sensor result is a screening flag, corroborating evidence, a trigger for follow-up sampling, or part of a formal compliance assessment. Without this hierarchy, even technically impressive sensor results may be politically unusable.

Limitations, Risks, and Areas Requiring Caution

The main risk is premature confidence. Across all sensor types, performance depends heavily on the measurement setting, quality of calibration, availability of reference data, and the way results are interpreted. Quantum gravimeters can improve local underground mapping, but measurements can be slow, sensitive to vibration, and difficult to interpret in complex geology or built environments. Magnetometers can detect magnetic anomalies, while buried infrastructure, vehicles, cables, and natural field variations can produce misleading signals. Quantum inertial systems can reduce navigational drift, yet integration with maps, classical sensors, and inspection procedures remains essential.

For nonlinear interferometers, laboratory spectroscopy is more mature than drone-mounted or open-path sensing, where alignment, humidity, background gases, sunlight, platform motion, and mixed samples can strongly affect performance. Detection limits for warfare-relevant chemicals should be established through realistic tests with safe simulants rather than inferred from ideal laboratory conditions. Compact lab-in-a-box systems are plausible future development goals, but do not yet reflect current capabilities. For biological applications, these sensors are best described as screening tools: they can help flag suspicious material but reliable agent identification still requires established genomic, immunological, culture-based, or mass-spectrometric methods.

A second risk is confidentiality. More capable field sensors may increase the amount of collateral information collected during an inspection. Verification organizations should therefore design quantum-sensor pilots alongside data-minimization rules: restricted spectral windows, pre-approved maps, on-site processing, tamper-evident logs, joint inspector–host review, and deletion or redaction rules for data outside the mandate. Such measures are a precondition for making new measurements politically acceptable.

A third risk is political and procedural rejection. Even a well-validated instrument may be refused if parties consider it too intrusive, insufficiently transparent, or outside the agreed mandate. Technical trials should therefore be accompanied by early consultation on the acceptability of equipment, data access, and rules governing decision-making.

Conclusions: Incremental Change, Not Revolution

Quantum sensing may eventually matter for verification because it could make certain measurements more mobile, selective, stable, and robust. Gravimetry may improve local underground mapping; magnetometry may support searches for metallic artifacts; inertial navigation could make inspections less dependent on external signals; and nonlinear interferometers could bring mid-infrared screening closer to field use. The evidence reviewed here does not yet show that these systems are ready for treaty inspections or consistently outperform mature alternatives. The main recommendations are pragmatic: fund a small number of comparative pilots focused on concrete verification questions, specify evidentiary thresholds, certify the data chain, and agree data-governance and equipment-approval rules before deployment. If those trials show a meaningful advantage, quantum sensing could become a useful addition to the verification toolkit.