Agentic AI is changing Cybersecurity in 2026

Agentic AI does not just answer questions. It takes action, connects to services, handles credentials and operates inside real environments. In this Security Strategist clip, Kevin Curran, Professor of Cybersecurity at Ulster University and IEEE Senior Member, joins host Trisha Pillay to explain why agentic AI changes cybersecurity assumptions and why security has to be designed into systems from the beginning. The conversation covers what AI is authorised to do once it is inside an environment, and why enterprises need least privilege, secure by design, privacy by design, input and output filtering, logging, minimal access and architectural constraints before agentic AI adoption runs ahead of control. Topics covered: • Agentic AI and cybersecurity • Secure by design • Privacy by design • Least privilege • AI agents and credentials • Input and output filtering • Logging and architectural controls Watch more EM360Tech cybersecurity conversations: https://em360tech.com/podcasts/how-agentic-ai-reshaping-cybersecurity #AgenticAI #Cybersecurity #AISecurity