Republic of Korea-hosted OSCE Asian Partners for Co-operation Group Meeting: REAIM Summit 2024 and Responsible application of Artificial Intelligence in the Military Domain
Statement by Dr Katie Mead, Australia’s Deputy Permanent Representative to the United Nations and International Organisations in Vienna
11 November 2024
Excellencies, colleagues,
I thank North Macedonia and the Republic of Korea for raising this timely and important issue.
I also thank our speakers for their insights and their continuing contributions in their respective fields.
Australia is pleased to be able to contribute to this discussion today and to bring an Indo-Pacific perspective to this forum. Our discussion comes at a critical time for the security of the OSCE area and the world.
Russia’s illegal and immoral invasion of Ukraine is a grave threat to regional and global security. And as we said during last week’s joint FSC-PC meeting, Australia is gravely concerned at the deployment of North Korean troops to Russia, and the DPRK’s continued transfer of weapons to Russia.
In these challenging circumstances, we commend the OSCE’s continued focus on the intersecting challenges impacting global security, including artificial intelligence in the military domain.
Australia recognises AI as a critical suite of technologies with potential to deliver significant benefits.
We are conscious that with potential benefits, the military application of AI may also generate new potential risks.
Australia is uplifting our existing governance, systems of control and procedures, to ensure military AI is in line with our commitments to responsible use. However, national action alone is not sufficient. Ensuring a shared international understanding of what it means to be a responsible AI actor in the military domain is crucial.
Australia welcomed the opportunity to participate in the second REAIM Summit in Seoul in September [9-10 September 2024].
Australia was pleased to join more than 60 countries in endorsing the REAIM Blueprint for Action.
Australia was also pleased to see the first resolution on artificial intelligence in the military domain adopted by UNGA First Committee last week [6 Nov]. We applaud the Republic of Korea and the Netherlands for this initiative.
We welcomed the inclusion of text on the possible challenges associated with AI in the military domain in regard to gender, race, age and other social aspects through bias in datasets or other algorithmic biases.
Multilateral and multi-stakeholder initiatives like REAIM provide the opportunity to develop norms, set future direction and share lessons learned, benefiting from expertise across all sectors.
Australia has and will continue to engage actively with the international agenda towards responsible approaches to AI in the military domain.
We look forward to continued collaboration with the OSCE, Participating States and Asian Partners for Co-operation, to address the challenges and opportunities posed by military use of AI.
I have a question for the panel:
We often see technologies, such as AI, reflect the biases that exist in the offline world. For example, researchers suggest that facial recognition software recognises male faces far more accurately than female faces. What can we as States do to ensure the development of these emerging technologies avoids or mitigates biases and does not replicate or amplify gender and other stereotypes?
Thank you.