Autonomy Under Law
Legal Guardrails for AI
* Skip to 23.40 mins for start of audio recording
Emerging Directions
Autonomous AI systems are proliferating. Effective Governance of AI systems is a global problem that no one has solved and is a major economic threat to Australia. What are we doing about it?
Scientists, engineers and legal experts from Nooriam, The UWA Data Institute, The University of Western Australia, Curtin University, the Pawsey Supercomputing Research Centre, LexChip, Stirling & Rose, and Murdoch are working together in the Autonomy Under Law Project to build empirical evidence as to how Australian built techno-legal infrastructure can reduce the threat of AI harms.
Over four months, the consortium will run extensive simulation testing on Pawsey’s supercomputing infrastructure, comparing AI agents using legal guardrails against AI without legal guardrails.
The Project will measure whether a legal harness can reduce agent harms. The results will be published in a joint technical report that will be shared with government and industry.
The launch will introduce the experiment, the legal harness architecture and the partners behind it, followed by an opportunity to meet the project team.
What happens next?
For us at the UWA Data Institute, this is exactly the kind of research collaboration we want to be part of, bringing together data science, AI, law, engineering, supercomputing, industry and government around a problem that is moving much faster than traditional governance frameworks.
The launch marks the beginning of a broader conversation around law, AI and the frameworks needed to guide emerging technologies.
As the Autonomy Under Law project progresses, we’ll be sharing further developments, research insights and opportunities to engage.
* Skip to 23.40 mins for start of audio recording
Full Gallery
Get in Touch
Our door is open for collaboration opportunities, membership enquiries and everything data science, machine learning, language models and AI. We encourage you to reach out to the team at the UWA Data Institute.