Reimagining Verification Under the Biological Weapons Convention in the Age of Synthetic Biology

The Biological Weapons Convention remains one of the great normative achievements of twentieth-century disarmament law. Its central command is spare and uncompromising: biological and toxin weapons must not be developed, produced, stockpiled, acquired or retained. The weaponisation of disease is placed beyond the pale of lawful state conduct. That achievement should not be understated. The BWC did not merely regulate a weapons category. It delegitimised it. Yet the Convention has always carried a structural defect. More than fifty years after its opening for signature, it still lacks a binding verification mechanism. That absence is not a technical footnote. It is the central institutional weakness of the biological disarmament regime.

The failed Ad Hoc Group negotiations, which collapsed in 2001, still cast a long shadow. Since then, states have disagreed over intrusiveness, confidentiality, export controls, biotechnology transfer, equality of obligations, and the risk that verification might become discriminatory in practice. Those objections remain politically real. A comprehensive protocol is not likely to appear by diplomatic magic. But the legal and scientific landscape has changed. The life sciences have not paused while states continue to debate institutional form. Gene editing, DNA synthesis, artificial intelligence, automated biofoundries, cloud laboratories and platformised biotechnology have transformed the field. They have expanded the frontiers of medicine, agriculture, diagnostics and public health. They have also made the boundary between peaceful and hostile biological activity harder to police.

That is the central problem. The BWC’s prohibition remains strong. Its compliance architecture does not. The correct response is not to reopen the Convention’s core prohibition. Article I remains fit for purpose precisely because it is drafted through the general-purpose criterion. It does not prohibit only named agents or fixed technologies. It prohibits biological agents and toxins of types and in quantities that have no justification for prophylactic, protective or other peaceful purposes. It is purpose-based. It is technology-neutral. It is capable of accommodating synthetic biology, AI-assisted design and future biological technologies not yet imagined.

The difficulty lies elsewhere. A purpose-based prohibition requires a purpose-sensitive verification system. The BWC does not yet have one. Traditional arms-control verification models cannot simply be transplanted into biology. Nuclear arms control is built around relatively visible signatures: fissile material, enrichment infrastructure, warhead design, delivery systems and test indicators. Chemical weapons verification is anchored in scheduled chemicals, industrial production pathways, declarations, inspections and destruction procedures. In both fields, the object of verification is often a material thing. Biology is different.

The relevant object may not be a stockpile or a facility. It may be an activity, a pattern, a design choice, a missing declaration, a classified programme, a sequence file, a procurement trail, a delivery experiment, or a laboratory protocol. The same pathogen may be used for vaccine research or hostile development. The same sequencing capacity may support public-health surveillance or reconstruction of a novel agent. The same gene-editing platform may serve agricultural resilience or immune-evasion research. The microscope may not disclose the offence. The legal question is context.

This is why the old binary — full verification protocol or nothing — is false. It misunderstands the task. Verification under the BWC should not be treated merely as inspection. It should be understood as a broader legal architecture of assurance: a system that generates information, tests declarations, compares conduct over time, assesses anomalies, reviews national implementation, and subjects high-risk activities to structured scrutiny. Such a system would not prove the absence of every prohibited biological programme. No serious verification system can promise that. Its more realistic task is different: to make non-compliance harder to conceal, peaceful conduct easier to demonstrate, and unresolved ambiguity harder to ignore.

That requires a layered model. The first layer is a stronger confidence-building measure regime. CBMs already perform a soft verification function. They require states to provide information on biodefence programmes, high-containment laboratories, unusual outbreaks, national implementation and past offensive programmes. Their present weakness is practical rather than conceptual. Participation is uneven. Submissions vary in quality. Information is often too thin to sustain serious comparative assessment. Too much material is deposited, archived and forgotten. That must change. CBMs should become more regular, granular, standardised and reviewable. Reporting on biodefence programmes and high-containment facilities should be more detailed. National implementation submissions should be treated as central, not peripheral. A declaration that no one reads is not assurance. It is paperwork. The second layer is rigorous domestic implementation. Article IV is often underread. It requires states not only to prohibit but also to prevent activities contrary to the Convention. Prevention is an active legal duty. It requires criminal law, licensing, biosafety rules, biosecurity frameworks, export controls, oversight of dual-use research, inspection powers, reporting duties, sanctions and institutional accountability.

This is where due diligence matters. A state cannot discharge its BWC obligations by passing a statute and hoping for the best. Hope is not a compliance mechanism. A credible domestic system creates the audit trail that international review requires. It shows who licenses, who inspects, who reports, who sanctions and who answers.

The third layer is focused peer review of high-risk facilities and programmes. Not all biological activity presents the same compliance risk. A teaching laboratory, a vaccine facility, a national pathogen archive, a BSL-4 laboratory and a military biodefence programme cannot sensibly be treated as equivalent. Equality of legal obligation does not require factual blindness.

Verification resources should be directed where risk and capability are highest. Voluntary peer review, transparency visits, compliance assessments, staff exchanges and models such as 3D BIO — declare, document and demonstrate — have particular value. They build operational familiarity before crisis. They test confidentiality rules. They allow states to understand one another’s systems in practice. They are not a substitute for binding verification. Properly framed, they are rehearsal for it.

The fourth layer is targeted reporting and anomaly assessment. The BWC needs a better way of dealing with warning signs before suspicion becomes accusation. Relevant indicators may include unusual combinations of gain-of-function work and aerosol testing, unexplained animal trials, secretive field activity, inconsistent declarations, unusual procurement, disinformation, or outbreak patterns that do not fit natural disease ecology.

Emerging technologies may assist. High-throughput sequencing, microbial forensics, biosurveillance, open-source intelligence and DNA synthesis screening can support directed inquiry. These tools will not produce perfect proof. They can, however, raise the cost of concealment and provide a basis for clarification under Article V or, in serious cases, an Article VI complaint to the Security Council.

The fifth layer is a permanent science-and-technology advisory mechanism. This is not administrative decoration. It is indispensable. Verification in the biological field cannot be static because biology is not static. CRISPR, AI-assisted design, DNA synthesis, laboratory automation, biofoundries and novel delivery systems alter both the risk landscape and the tools available to manage it. A treaty that cannot learn will eventually fail to govern.

A standing S&T mechanism would help states identify emerging risks, distinguish serious misuse pathways from speculative anxieties, assess new monitoring tools, and update indicators of concern. It would also reduce the risk of policy being driven by theatrical worst-case scenarios rather than actual laboratory practice.

That distinction matters. Synthetic biology should not be treated as a magic wand for mass-casualty biological weapons. The idea that cheap DNA synthesis or gene editing has made catastrophic misuse easy is often overstated. It ignores tacit knowledge, institutional infrastructure, formulation challenges, delivery problems and the weaponisation step itself. The point is not to minimise risk. It is to locate it accurately. Verification built on fantasy will fail. Verification built on realistic pathways to misuse may succeed.

The political objection to incrementalism is serious. Many states, particularly within the Non-Aligned Movement and parts of the Asia-Pacific, continue to insist on a non-discriminatory, balanced, legally binding verification mechanism. Their concern is legitimate. Voluntary and modular tools may entrench asymmetry. Powerful states may influence standards without accepting equivalent duties. Incremental measures may become an excuse for avoiding a protocol indefinitely. Those risks must be met directly. Incremental measures should be framed as building blocks, not substitutes. Reformed CBMs, domestic implementation review, peer visits, anomaly assessment and science advice should be treated as institutional testbeds for future legal codification. They should preserve, not abandon, the long-term objective of binding verification.

The Ninth Review Conference and the Working Group on strengthening the Convention provide the present institutional opening. The Working Group’s mandate covers compliance and verification, CBMs, national implementation, science and technology review, international cooperation and assistance. That breadth is not accidental. It reflects the reality that BWC verification cannot be solved by inspection alone. It must be built across legal, technical, scientific and institutional domains. The deeper legal basis is good faith. Pacta sunt servanda requires treaties to be performed in good faith. Under the BWC, good faith is not exhausted by saying, “we have no biological weapons.” It requires credible prevention, transparency, cooperation and assurance. It requires states not to hide behind ambiguity when the object and purpose of the Convention demand seriousness.

Article X must also be kept in view. Verification must not strangle peaceful biological cooperation. The BWC is not an anti-science instrument. It protects the peaceful use of biology while prohibiting its weaponisation. A credible verification architecture must therefore be firm but proportionate. It must protect legitimate research, public health and development. It must not treat every laboratory as a crime scene. This is the balance that a mature BWC regime must strike. Too little scrutiny invites mistrust. Too much intrusion damages science and encourages resistance. The correct aim is targeted scrutiny of activities that are objectively high-risk, combined with clear protection for peaceful research.

The Convention’s future will not be secured by rhetoric. The proposition that biological weapons are unlawful is settled. The harder task is institutional: to build a system capable of showing that the norm is being observed. That is why the verification debate must move beyond nostalgia for a failed protocol and beyond resignation to permanent weakness. The BWC does not need to become the Chemical Weapons Convention in borrowed clothes. It needs to become a stronger version of itself. A credible future model should rest on five connected elements: stronger CBMs; serious domestic implementation; focused peer review; targeted anomaly assessment; and a standing science-and-technology advisory function. None of these elements is sufficient alone. Together, they can create a layered architecture of assurance suited to a dispersed, dual-use and rapidly evolving biological domain.

The age of synthetic biology does not render the BWC obsolete. It makes the Convention’s general-purpose criterion more important. But it also exposes the inadequacy of compliance machinery designed for an earlier scientific era. The aim should now be clear. Peaceful conduct must become easier to demonstrate. Suspicious conduct must become harder to conceal. Unresolved ambiguity must become harder to tolerate. That is the proper measure of verification in the biological field. Not perfect certainty. Not institutional theatre. Not inspection for its own sake. Structured assurance. The BWC’s prohibition remains sound. The machinery around it must now become serious.