January 2025 Issue

HSP AI-CBW Chronology

Since 2022, HSP has been chronicling policy documents, statements, and news items which are of relevance to the convergence of AI technology in chemical & biological security.

This January 2025 edition includes a number of significant updates, including:

5 September 2024 - EU AI Treaty Opens for Signatories
The Council of Europe Framework Convention on Artificial Intelligence, Human Rights, Democracy, and the Rule of Law was opened for signatories. Council of Europe Secretary General Marija Pejčinović Burić commented on the purpose of the treaty:

“The Framework Convention is designed to ensure that AI upholds our standards. It is a text developed through an open and inclusive approach, benefiting from multiple expert perspectives.”

8-11 October 2024 - OPCW 107th Executive Council
The Executive Council holds its 107th Session at OPCW headquarters in The Hague. For the first time, a regular item focusing on emerging technologies has been added to the agenda, as Agenda Item 10.

12 December - “Tripwire Capabilities for AI”
The Carnegie Endowment for International Peace publishes a report by Holden Karnofsky titled “A Sketch of Potential Tripwire Capabilities for AI.” The report proposes the establishment of specific “tripwire capabilities” to identify and mitigate risks associated with advanced AI systems before they reach critical thresholds.

The full chronology can be found here.

CWC Statements Dataset

A dataset of AI-related statements made at CWC meetings

This dataset is being compiled by Dr Alexander Ghionis as part of a wider effort towards understanding topics under discussion in CBW forums. The dataset extracts statements made in sessions and meetings of the policy making organs of the Organisation for the Prohibition of Chemical Weapons (OPCW) that refer or allude to Artificial Intelligence. Each statement is categorised by the meeting at which it was made, the actor it was made by, and the agenda item under which it was made.

The full dataset can be found here.

The dataset is a work in progress. It is important to note that not all statements are made available on the OPCW website. A number are also not made available in English. The dataset therefore is partial and representative, rather than conclusive. The dataset is being updated as documents become available. In the coming months we aim to expand the dataset to include references made within the Biological Weapons Convention.

Past as prologue, future as framework

The importance of history in predicting and responding to technological change

Brett Edwards is Senior Lecturer at the University of Bath, specialising in ethics and governance of technology for defence, security, and resilience. He also hosts and produces the Poisons & Pestilence podcast which explores the history of chemical and biological weapons.


I feel very lucky to have been asked to contribute to this first issue. ‘Interface’ is an apt title for the series. It speaks to open-ended interactions between different knowledge communities. Such dialogue really is important. It takes much longer for individuals to develop new technical understandings than it does to build new forms of reliable relationships with those that already have them.

In that spirit, then, I wanted to use this opportunity to emphasise the value of historical chronologies, and indeed historical approaches more generally. These are essential components of the intellectual work which underpin attempts to characterise the contemporary political and technological landscape, anticipate developments, and better conceptualise the challenges at hand. Historical thinking, as well as rigorous historical empiricism, are vital in providing perspective on contemporary events and checking over confidence when making predictions about the future.

A case in point is the historical role of technological change in the evolution of chemical and biological weapons, and the global prohibition against them. While we often use technological developments as a wristwatch to track historical progress (a cliché has itself become ironically dated), technologies are not some external factor which have driven humankind along since the pre-historic area. Technologies emerge from a time and place, and they shape that time and place.

This creates a real problem when trying to make decisions about the interface between new technological developments and existing institution. In response, there is a tendency to be reductive when considering technological change over time. Something which tends to manifest in two distinct, but similar, mappings of the future.

First, we tend to consider the impact of technological change on a largely static socio-cultural landscape. For example, how will today’s cutting-edge biotechnologies develop further and be democratised; and how might these undermine our existing structures and policies of non-proliferation? In this way, we extrapolate the future of a technology to find its impacts on today’s socio-cultural landscape.

The second is to consider how changes in the broader socio-cultural landscape, mediated usually by substantial policy shifts, might affect current technological trajectories. For example, will regulation in particular technological domains stifle innovation and change future technologies? Here we are extrapolating potential future policy changes and examining impacts on today’s technologies and innovation ecosystems.

In either of these cases, we are attempting to reduce the complexity of potential futures by examining changes in one landscape while holding the other static. This is logical, but often ignores the interactions between the technology and society. While regulation may reduce scope for some innovations, the history of CBW (and indeed broader technology) shows us that technologies often develop around regulatory barriers; necessity is said to be the mother of invention after all. Similarly, technological changes can instigate significant reshaping of societal relationships and structures. In essence, changes in either the technological or socio-cultural landscapes necessarily effect change in the other – the two co-evolve.

More fundamentally however, we need to appreciate not only that technology and society are in a co-productive relationship, but that this is only part of the bigger historical picture when dealing with the niche which chemical and biological weapons have evolved to fill.

On New Technology, History and Chronologies

There have been several key drivers which have shaped the history of chemical and biological weapon development and use. Most significantly, developments in the character of warfare and violence, broader shifts in the technological base, as well as evolution in how states and societies identify, characterise and respond to threats.

The most significant developmental periods in the history of this area involve a confluence between these shifts. For example: the emergence of chemical weapons in WW1 was not something which was shaped exclusively by the emergence of a large chemical industry; nor can it be understood to have emerged exclusively because of the specific demands created upon states by this conflict; nor did the use of these weapons occur in a moral vacuum, after some invisible watershed moment was passed. Instead, the history of the emergence and use of chemical weapons was a complex and iterative process. Novel programmes and doctrines of offense and defence married emergent technical capabilities with cultural norms within both state militaries and those societies more broadly. The same might be said in relation to the emergence of the gargantuan Soviet biological warfare programme of the late Cold War era. This was a product of the byzantine command economy and security architecture, in as much as it was a product of militaristic and industrial drivers. And we can of course detect related cultural and institutional dimensions in the emergence of the US biodefense complex in the early 2000’s.

From this historical perspective, addressing questions of why and when, there is a tendency to focus on the examination of key junctures in which new strategies and capabilities emerged. These moments centre on broader political and material shifts, as well as the work of individuals and institutions who seek to bring about changes incrementally in the context of this complex and evolving landscape. And yet, when we ask ourselves about contemporary emerging developments, and how they might shape the contemporary threat landscape, it is often the case that we tend to forget the convergent and contingent character of change. Instead, for example, we might focus on hypotheticals related to impact of technological change on non-proliferation systems. Or perhaps consider how broader trends in terrorism might increase the appeal of CBW in discrete, linear moves.

And this is where historical thinking is of value. On the one hand, it reminds us that every new development happens in broad historical moments. On the other, it helps us consider the complex interactions between developments in innovation, conflict, and society more broadly.

To give a concrete contemporary example: in the field of AI, there has been a proliferation of hypothetical misuse scenarios in the academic and policy literature. These tend to focus on either the proliferation of capabilities to the masses, or else the potential utility of such developments to well-resourced terrorist groups, or state level programmes. In such cases, in addition to mapping the future in the two ways discussed above, analysis tends to extrapolate from the rather shallow histories of bioterrorism, as well as increasingly anachronistic conceptions of the military utility of CBW which solidified during the Cold War era.

I would suggest perhaps, that within much of this contemporary analysis there is a tendency to draw the wrong type of insights from history. Historical chronicles of technological and social change are important then, when it comes to asking three types of question today. First, they help us to better capture the evolution of the motivations of those that would pursue chemical and biological weapons, as well as those that would seek to maintain the prohibition. Second, they help structure thinking about ‘how’ questions, scaffolding discussions of the timescales and complex processes through which changes in policy, technology, and doctrine co-evolve. Finally, and perhaps most profoundly, they help with the ‘what’ questions - what do we actually seek to control when we seek to manage proliferation? What is included, what is left out, and why? The answers to these questions, in the most general sense, have varied throughout the history of chemical and biological non-proliferation and will continue to do so.

Chronologies of key historical junctures, then, remain essential, if often under-utilised resources of where we have been, where we are, and where we might be are heading.

Inaugural Issue

Issue No 1 - January 2025

Welcome to the first issue of INTERFACE, a newsletter focused on the intersection of artificial intelligence (AI) and chemical and biological weapons (CBW). This initiative was inspired by discussions with colleagues and insights gathered at various events, reflecting a shared desire for a clearer perspective on the growing discourse around AI in our field.

If you would like to to receive future updates of INTERFACE, you can sign up here.


This inaugural issue includes:

Past as prologue, future as framework - Guest Editorial by Brett Edwards

Updated AI-CBW chronology

A new dataset on AI-relevant statements made at CWC meetings



INTERFACE aims to provide an overview of relevant developments, publications, and brief analytical comments on AI’s potential implications for CBW governance. Our goal is to contribute to ongoing discussions by helping to capture what’s being said, by whom, and what’s happening in this rapidly evolving area. By doing so, we hope to provide a foundation that can support and facilitate evidence-based work and analyses by professionals in the field. We hope this format proves useful for busy professionals in disarmament, non-proliferation, and emerging technologies.

Looking ahead, we welcome short contributions (up to 1000 words) from readers for future issues, as well as suggestions for relevant events or publications to feature. Your input will be valuable in shaping this resource to serve our community’s needs.

INTERFACE is an experiment, one small tool among many in this field. We see it as a modest contribution to collective efforts in understanding the AI-CBW nexus. We look forward to your feedback as we continue to work together to better ground, interpret, and respond to the implications of AI for the CBW regimes.

Welcome to INTERFACE

A regular update from HSP on all things AI-CBW

Providing regular updates on our AI-CBW chronology, curated datasets, and guest editorials, INTERFACE is a central convening space for those who want to keep up with developments in the convergence between Artificial Intelligence and Chemical & Biological Security.

See below for current and previous editions.