Johanna Weaver

Who Determines What Safe Is? Gender, Diversity and Power in the Future of AI

AI safety is often framed as a technical challenge, but the way risks are identified, prioritised and governed is deeply shaped by who is in the room. This panel will explore why gender, diversity and lived experience matter to the future of AI safety, and what is lost when the field is dominated by narrow perspectives. Speakers will consider how AI systems can reproduce or deepen existing inequities, how diverse expertise can strengthen risk identification and governance, and what practical steps are needed to make AI safety a more inclusive and publicly accountable field.

A Scientist, a Sceptic and a Diplomat: What Should AI Safety Prioritise Now?

AI safety is an increasingly crowded and contested field, with different communities prioritising different risks, timelines and interventions. Some argue that we need to focus on long-term existential risk, while others that argue we need to focus on immediate harms. This panel brings together distinct perspectives. Expect robust but respectful discussion. We will test where there is genuine consensus, where disagreement is productive, and how Australia should approach prioritisation in a field where the stakes are high and the evidence base is continually evolving.

From Risk to Response: Mapping AI Harms and Unintended Consequences

This workshop is the final session in a four-part series delivered in the lead-up to the AI Safety Forum, bringing together cross-sector participants to examine what is needed for Australia to adopt AI safely, confidently and at scale. Building on previous roundtables, the session will focus on the opportunities, barriers and emerging harms associated with wider AI use across industry, government and society. It will ask how we can better identify future risks, anticipate unintended consequences, and prioritise the issues most likely to shape Australia’s AI readiness. Participants will work through structured prompts on current and future uses of AI, including how emerging harms or behaviours can be recognised, assessed and managed without defaulting too quickly to narrow regulatory or technical solutions. The workshop is designed to support a practical, cross-sector conversation about where Australia should focus its effort to build trust, capability and responsible adoption.