Gradient Institute is an independent Australian nonprofit research organisation advancing safe and responsible AI through rigorous science-based research, practical guidance, and policy engagement.
Gradient Institute is an independent Australian nonprofit research organisation advancing safe and responsible AI through rigorous science-based research, practical guidance, and policy engagement.
People
Updated 05/18/26Director of AI Practice
Gradient Institute Fellow
Principal AI Specialist
Fellow
Funding Details
Updated 05/18/26- Annual Budget
- -
- Current Runway
- -
- Funding Goal
- -
- Funding Raised to Date
- -
Org Details
Updated 05/18/26Gradient Institute is an independent nonprofit research organisation based in Sydney, Australia, founded in 2019 by Bill Simpson-Young and Dr. Tiberio Caetano. The institute emerged from a founding partnership between the University of Sydney, IAG (Australia's largest general insurer), and CSIRO's Data61, reflecting an ambition to bridge academic research, industry practice, and public policy on AI ethics and safety. The organisation's work spans four main areas: capability and safety research investigating how AI systems behave and where they fail; sociotechnical impact research examining how AI shapes decision-making and institutions; policy and governance work translating technical insight into practical regulatory guidance; and applied research and education helping organisations adopt AI responsibly. The institute positions its independence — answering only to its mission rather than commercial or political interests — as a core differentiator that enables trusted, impartial analysis. Key projects and achievements include co-developing the Guidance for AI Adoption framework with Australia's Department of Industry, Science and Resources and the National AI Centre; publishing a Multi-Agent Risk Analysis report on emergent dangers in autonomous AI systems; delivering a Responsible AI Support Program for Australian nonprofits with Google.org support; contributing to Australia's National Framework for AI Assurance in Government; and hosting the inaugural Australian AI Safety Forum in November 2024 in Sydney. The institute has also collaborated with the Minderoo Foundation on open-source tools for AI impact assessment and received research funding from the Telstra Foundation and Cadent. As of early 2026, the core team comprises nine staff members, including Bill Simpson-Young (Chief Executive), Tiberio Caetano (Chief Scientist), Simon O'Callaghan (Head of Technical AI Governance), Ali Akbari (Director of AI Practice), and Alistair Reid (Head of Research Engineering). The team is supported by twelve fellows drawn from universities and industry, two non-executive board members, and an advisory board that includes the Australian Information Commissioner.
Theory of Change
Updated 05/18/26Gradient Institute believes that independently produced, rigorous scientific research about AI systems — their capabilities, limitations, failure modes, and societal impacts — is a critical input to better AI governance and safer AI deployment. By translating complex technical evidence into practical guidance, the institute aims to improve the quality of decisions made by governments, regulators, and organisations about how AI is built, procured, and governed. Better-informed policy and governance is expected to reduce the likelihood of harmful AI deployments, embed accountability and transparency into AI systems at scale, and build societal capacity to manage AI risks over time. The independence of the institute from commercial and political pressures is viewed as essential to credibility and impact.
Grants Received
Updated 05/18/26Projects– no linked projects
Updated 05/18/26Discussion
No comments yet. Be the first to share your thoughts.