Activate existing biosecurity powers to address AI-enabled risks
Letter to Minister for Agriculture, Fisheries and Forestry
Sent 4 May 2026
Dear Minister,
Australia has one of the world's most rigorous biosecurity frameworks, and a new gap has emerged that it is well-placed to close. AI and synthetic biology now pose increasing and novel risks that require immediate action from your department.
Synthetic nucleic acids can be used to develop vaccines, improve crops, and advance medical research, but also to engineer or construct pathogens. Australians can order custom nucleic acids from overseas providers. Previously, the key constraint in creating a biological weapon from synthetic nucleic acids was the knowledge required — expertise that was rare and hard to acquire. That has changed.
The International AI Safety Report 2026 confirms that general-purpose AI systems can now provide expert-level guidance on biological and chemical weapons development, including detailed laboratory instructions.1 One study found an AI model outperforming 94% of domain experts at troubleshooting virology protocols.2 Specific AI systems (biological foundation models) can generate genomes for viruses (bacteriophages) not previously seen in nature.3 Related techniques have already been used to design protein variants that evade the human immune system.4
Current AI practices do not adequately address this risk: only 3% of 375 biological AI tools surveyed have any safeguards,5 and where they exist, ways to circumvent the restrictions (jailbreaks) are common.6
A bad actor with AI guidance still needs the materials to act, and imported synthetic nucleic acids are a physical chokepoint. Australia's Biosecurity Import Conditions system (BICON) gives the Director of Biosecurity the power to impose conditions on all such imports — immediately, without new legislation. Members of the International Gene Synthesis Consortium, including leading providers such as Twist Bioscience, IDT, and GenScript, already screen orders for sequences of concern voluntarily. Yet, Australia's current import permit system does not require that synthetic nucleic acids be sourced from providers that screen, and bad actors could exploit that.
Australia has committed to reducing biosecurity risks to a very low level.7 The science now provides reasonable grounds for biosecurity officers to conclude that synthetic nucleic acids imported without safety screening pose an unacceptable risk under the Biosecurity Act 2015.
The US,8 New Zealand,9 the EU,10 and the UK11 are each taking steps to address screening requirements. Australia can do the same, and without new legislation.
The National AI Plan says that regulators are responsible for identifying and addressing AI-related harms within their domains.12 We call on you to act on that responsibility.
We recommend that:
- A screening condition be applied to all BICON import permits for synthetic nucleic acids. Permits should require that synthetic nucleic acids be sourced from providers that screen orders for sequences of concern and, where sequences of concern are identified, verify customer identity and legitimacy. Appropriate mechanisms should be in place to verify provider compliance.
- Applicants for higher-risk synthetic nucleic acid import permits are prioritised for the statutory fit and proper person test13 — a government background check that goes beyond provider screening — with the department publishing clear criteria for when and how the test is applied.
- The department reviews the adequacy of the regime within its broader context, in consultation with relevant Commonwealth agencies, including the new AI Safety Institute, civil society, industry, and academia. The review should address progress made in AI and biotechnology as well as emerging issues such as domestic production of synthetic nucleic acids, control on benchtop DNA synthesisers, and mechanisms for verifying provider compliance.
These steps should be accompanied by appropriate transparency to signal to the public, the global community, and potential bad actors that Australia is taking action.
Implementing these measures is low cost.14 Screening is conducted digitally by the synthesis provider before dispatch, imposing no additional burden on Australian researchers, businesses, or the regulator. Free screening tools are available.15,16 Major providers already screen voluntarily — this change targets the gap, not the norm.
This is an opportunity for Australia to address domestic risks and build on its world-leading biosecurity framework. As a founding member of the Australia Group, Australia has a track record of international biosecurity leadership. Mandating gene synthesis screening would be a significant step toward establishing a global norm and addressing AI-biosecurity risk.
Yours faithfully
The undersigned
Individual Signatories
Note: Signatories endorse only the core letter text. Footnotes and additional content may not represent their views.
Dr. Cassidy NelsonDPhil MBBS MPH
Centre for Long-Term Resilience
Director of Biosecurity Policy
Dr. Brendan Walker-MunroPhD
Southern Cross University
Associate Professor
Managing Editor of Routledge International Handbook of Research Security
Janet Egan
Center for a New American Security
Senior Fellow and Deputy Director
Ms Kate ChaneyMP
Federal Member for Curtin
Dr Lotti Tajouri
Bond University and Murdoch University
Associate Professor
Dr. Alexander SaeriPhD
MIT FutureTech
Director, AI Risk Initiative
Dr. Sarah Winthrope
Brown University Pandemic Center
Visiting Fellow
Mr Soroush Pour
Harmony Intelligence
CEO
Fmr Head of Technology at Vow (world leading biotech firm)
A/Prof David HeslopFAFOEM FRACGP MBBS PhD MPH BSc(Adv) Hons 1 MAICD
University of New South Wales
Associate Professor
Prof Patrick F Walsh
Charles Sturt University
Professor, Intelligence and Security Studies
Mr Rumtin Sepasspour
Global Shield
Director of Policy and Strategy
Dr. Michael Noetel
University of Queensland
Associate Professor
Dr. David Manheim
Technion - Israel Institute of Technology / Alter.org.il
Director of Research and Policy
ISO/IEC JTC1 SC42 Expert
Dr. Ryan KiddPhD
MATS Research
Co-Executive Director
Co-Founder, London Initiative for Safe AI
Prof Nick Wilson
University of Otago (New Zealand)
Research Professor
Dr. Peter Slattery
Massachusetts Institute of Technology
Research Scientist
Dane Sherburn
p-zero research
CEO
Formerly Preparedness Team (Contractor) at OpenAI
Scott Weathers
Americans for Responsible Innovation
Associate Director of Government Affairs
Dr Sam BuckberryPhD
The Kids Institute Australia, Australian National University
Head, Epigenetics
Prof David Balding
University of Melbourne
Honorary Professor of Statistical Genetics
Dr Adam Kamradt-ScottPhD
The Fletcher School of Law and Diplomacy, Tufts University
Cummings Foundation Associate Professor of One Health Diplomacy
Mr Michael Clark
Cytophenix
Director
Bill Simpson-Young
Gradient Institute
Chief Executive
Dr Piers Millett
International Biosecurity & Biosafety Initiative for Science (IBBIS)
Executive Director
Mr Zac Hatfield-Dodds
Anthropic
Member of Technical Staff
Biosecurity lead, Claude Opus 3
Dr Duncan PurvisPhD
Volunteer with Australians for Pandemic Prevention
Dr Toby Ord
Oxford University
Senior Researcher
Author of The Precipice: Existential Risk and the Future of Humanity
Karl Berzins
FAR.AI
Co-founder & President
Mr Devon Whittle
Global Shield Australia
Australia Director
Mr Jay Bailey
Arcadia Impact
Head of Technology and Standards
Former UK AISI Research Engineer
Keltan O'Shea
The Machine Intelligence Research Institute (MIRI)
Dr Jamie Freestone
AI safety researcher
Michael Kerrison
AI Safety Australia & New Zealand
Executive Director
David Conrad
Talos Network
Managing Director
Luke Freeman
Good Ancestors
COO
Mr Greg Sadler
Good Ancestors
CEO
Dr Saskia Popescu
Global Health Security Network/ RAND
CEO/Policy Researcher
John Pane
Electronic Frontiers Australia Inc.
Chair
Dr Kun Zhao
MIT FutureTech
Senior Researcher, AI Risk Initiative
Chris Leong
Sydney AI Safety Fellowship
Lead Organiser
Mr Yanni Kyriacos
Technical Alignment Research Accelerator
Director
Dan Braun
Goodfire
Mr Matt Fisher
Arcadia Impact
Senior AI Evaluations Engineer
Nathan Sherburn
Effective Altruism Australia
Chief Technology Officer
Dr Tim Seelig
University of Queensland
Adj. Assoc. Professor
Peter Horniak
PauseAI Australia
Director
Noel Lim
Anika Legal
CEO
2025 Victorian of the Year
Mr Martin Veron
University of Queensland
Doctoral Candidate
Dr Sam CogginsPhD
Australian National University
AI Risk Governance Researcher
Hugo Lyons Keenan
The University of Melbourne
ML PhD Student
Supporting Organisations
12 organisations have signed this letter
AI Safety ANZ
Signed by:
- Ms Emma Humphrey — New Zealand Community Lead and Ecosystem Builder
Atinar Pty Ltd
Signed by:
- Mr Ramakrishnan Veeramony MBA, MAICD — Managing Director
Beacon Institute for Global Catastrophic Risk
Signed by:
- Pip Foweraker — CEO
Center for Existential Safety
Signed by:
- James Norris — Executive Director
Electronic Frontiers Australia Inc.
Signed by:
- John Pane — Chair
Global Shield Australia
Signed by:
- Mr Devon Whittle — Australia Director
Good Ancestors
Signed by:
- Luke Freeman — COO
Melbourne Security Forum
Signed by:
- Andrew Harris — Co-Founder
PauseAI Australia
Signed by:
- Peter Horniak — Director
Sydney AI Safety Fellowship
Signed by:
- Chris Leong — Lead Organiser
Technical Alignment Research Accelerator
Signed by:
- Mr Yanni Kyriacos — Director
Uniting Church in Australia, Synod of Victoria and Tasmania
Signed by:
- Dr Mark Zirnsak — Senior Social Justice Advocate