Conference Description
Key Takeaways
- Australian edition of the US-based Unprompted conference, focused on AI-driven cybersecurity research
- Covers vulnerability research, exploit development, red and blue teaming, and malware analysis using AI
- Features insights from engineers and researchers at frontier AI laboratories
- Designed for security practitioners, AI researchers, technical leads, and engineering professionals
- Two-day single-track programme held in Sydney with hybrid attendance options
Introduction
[un]prompted.au is an Australian cybersecurity research conference examining the integration of artificial intelligence into security operations and research. Taking place in Sydney on 18–19 September 2026, the event brings together security practitioners, AI researchers, and engineers working at the intersection of these rapidly converging fields. As organisations increasingly deploy AI systems to scale threat detection, automate vulnerability discovery, and accelerate incident response, the conference addresses both the opportunities and security implications of these technologies.
About This Event
[un]prompted.au serves as the Australian edition of the US-based Unprompted conference, establishing a regional forum for AI security research. The two-day programme follows a single-track format at ILUMINA in Sydney, with virtual tickets available for remote participants. The conference is research- and practitioner-led, prioritising technical depth over commercial presentations.
Sponsors supporting the event include Google, Vigilant Labs, Catalyst Security, Dataflow Security, Epsilon Security, Elttam, and Calif, with Risky Business serving as media sponsor.
AI Integration Across Security Workflows
The programme addresses how AI is reshaping core security disciplines. Vulnerability researchers are using machine learning models to identify potential weaknesses in codebases at speeds impossible through manual review alone. Exploit developers are exploring how AI can assist in understanding complex attack surfaces and generating proof-of-concept code. Red and blue teams are incorporating AI into adversarial simulations and defensive monitoring, while malware analysts are leveraging automated classification and behavioural analysis.
These applications represent a fundamental shift in how security work is conducted. Rather than replacing human expertise, AI tools are extending the capabilities of skilled practitioners, allowing them to focus on higher-order analysis while automated systems handle pattern recognition and initial triage.
Perspectives from Frontier AI Laboratories
Beyond applied security, the conference features contributions from engineers and researchers working at frontier AI laboratories. These sessions examine the design, capabilities, and security implications of advanced AI systems themselves. Topics include automated research workflows, advances in inference and optimisation techniques, and the study of model behaviour.
Understanding how large-scale AI systems operate is increasingly relevant to security professionals. As these models become more capable, questions around their potential misuse, the security of their training pipelines, and their behaviour under adversarial conditions become critical areas of inquiry.
Industry Context
The convergence of AI and cybersecurity reflects broader industry pressures. Security teams face expanding attack surfaces, sophisticated threat actors, and a persistent shortage of skilled professionals. AI offers a path to scaling defensive capabilities, but it also introduces new risks. Adversaries are adopting the same technologies to automate reconnaissance, craft convincing phishing campaigns, and develop novel attack techniques.
This dual-use nature of AI in security makes events like [un]prompted.au particularly timely. Practitioners need to understand both how to deploy these tools effectively and how to defend against their misuse. The conference aims to foster a community equipped to navigate this evolving landscape, with plans to expand into governance and infrastructure topics as the field matures.
Who Should Attend
The conference is designed for professionals actively working in or transitioning into AI-augmented security roles. This includes vulnerability researchers, exploit developers, red and blue team operators, and malware analysts seeking to integrate AI into their workflows. AI researchers and engineers interested in security applications will find relevant technical content, as will technical leads, security engineers, and research scientists from cybersecurity firms, AI laboratories, and research institutions. Senior leaders such as CISOs and CTOs evaluating AI security strategies may also benefit from the programme’s coverage of emerging capabilities and challenges.