Topic: When the Attacker Is the Algorithm (AI Incidents and the Controls That Stop Them)
Abstract: Artificial intelligence does not replace traditional cybersecurity risk. It changes the speed, scale, credibility, and attack surface of that risk. The objective is not to block AI adoption. It is to make AI use defensible, controlled, auditable, and resilient. Join this Cyber Talks session to understand how organizations that inventory AI, control data flows, limit authority, test aggressively, and preserve
accountability can capture AI’s benefits while reducing the likelihood that it becomes the organization’s fastest path to a cybersecurity incident.
Key Takeaways:
- A general understanding of AI governance and ethics.
- Understanding what actions can AI agents take today, and what approvals, limits, and logs constrain them?
- Reconstructing an AI-related incident well enough to determine what data the system accessed, what sources influenced it, and what actions it performed.
Speaker:
Wes Knight, CISO at Needling Worldwide
Bio: Wes Knight brings more than four decades of information security experience, beginning his career at IBM in 1979. In 1996, he launched his own network security consulting and training practice before joining the Georgia Department of Revenue in 2008, where he served as Chief Information Security Officer, leading the agency’s Office of Information Security until his retirement in 2020.
Wes is currently CISO and Vice President of Government Accounts at Needling Worldwide, LLC, where he leads audits and assessments against ISO 27001, ISO 42001, CMMC, and NIST standards, and provides virtual CISO (vCISO) services. A frequent cybersecurity speaker, he is known for translating complex security and governance challenges into practical guidance for security leaders.
