Cyber-security needs to be embedded into the design and delivery of AI in order to protect society – not treated as an ‘afterthought’ - a leading expert will tell a conference in Belfast today (Wed 30 Sep).
Professor Maire O’Neill, who leads research at the Centre for Secure Information Technologies (CSIT) at Queen’s University Belfast, will give the keynote at the two-day CSIT Summit – an annual gathering of industry leaders, government officials and academics.
The theme of this year’s conference is ‘Cyber Security in the AI Era’ and discussions will range from immediate and practical solutions for enhancing security of AI to more philosophical questions such as whether rogue AI elements could destroy humanity.
READ MORE: New report identifies quantum opportunities for Northern Ireland
The Summit will take a considered view on the rapidly evolving threat landscape, including lessons to be learned from recent incidents involving AI-enabled systems and autonomous agents.
Prof O’Neill said:
“As AI systems become more capable and autonomous, security cannot be treated as an afterthought. It must be embedded into the design, development and deployment of these technologies.
“Rather than asking simply whether AI will transform society, the challenge is increasingly becoming how that transformation can be achieved securely, responsibly and with appropriate safeguards in place.
“The next generation of AI security will require greater transparency for users, improved resilience and availability, and a broader understanding of the potential consequences of deploying increasingly autonomous systems.”
She also said that while the public debate and speculation around AI has grown exponentially in recent years, the security of AI has been a long-term, key research focus for CSIT at Queen’s, as well as research into the use of AI to improve cyber security performance.
Prof O’Neill added:
“AI is already transforming sectors ranging from financial technology and healthcare to manufacturing and critical infrastructure. However, the growing use of Large Language Models, Agentic AI and Embodied AI is also creating new attack surfaces and security challenges.
“CSIT has been at the forefront of cybersecurity research for almost two decades, including the development of AI during that time period.
“Critically, CSIT is also home to the Cyber-AI Hub, launched in 2023 as a collaboration between government, industry and academia, focused on advancing research and innovation in AI security.
“The most recent and future challenges coming from AI will require a very close alliance between research, the technology sector and government and our conference this week will play a key role in knowledge-exchange, showcasing solutions-focussed collaboration.”
READ MORE: Empower Girls Returns for Fourth Year as Organisers Aim to Inspire More Young Women Than Ever Before
John Sotiropoulos, Founder of Deep Cyber; OWASP GenAI Security Project Board Director and Agentic Security Initiative Co-Lead, added:
“We live in profoundly changing times. OWASP, ETSI and the UK’s AI Cyber Security Code of Practice provide practical foundations for securing AI, but recent incidents show how agents with advanced intelligence and autonomy can pursue legitimate goals through harmful, unauthorised actions. This creates a new frontier to secure at machine speed.
“Organisations like CSIT are actively advancing research to address this and we need to collaborate across industry and academia. The OWASP GenAI Security Project’s Agentic Research Council will enhance our collaboration with thought-leading organisations like CSIT and help connect research with practical safeguards.”
Read the Summer 2026 edition free online →
Stay connected with NI's tech community:
