Organizations and individuals are feeding sensitive corporate data, source code, internal documents, credentials, and security decisions into AI systems without fully understanding the consequences. At the same time, many organizations are blindly relying on AI-generated code, AI security advice, vibe-coded applications, and automated decisions without realizing how often these systems hallucinate, fail, leak information, or confidently make terrible choices.
At the same time, threat actors are having a field day. Criminal groups are using AI to scale phishing, fraud, impersonation, romance scams, malware development, reconnaissance, and even penetration testing operations faster than most defenders can react. Some of the examples we will walk through show AI acting less like a tool and more like an accomplice that helps bad actors refine ideas, automate abuse, and remove skill barriers that used to slow criminals down.
As usual, this is not a theoretical AI hype talk. We will look at practical real-world examples, attacker workflows, operational failures, and the uncomfortable reality that many organizations are already giving AI enough trust and access to become part of their attack surface.
Watch on BrightTALK
