Defending Against Hidden Risks Of AI Tools In The Workplace

The balancing act of AI adoption acceleration is unlocking customer productivity gains without exposing organizations to new cybersecurity risks. Rob Allen, chief product officer at ThreatLocker said AI is lowering the barrier of entry for attackers while introducing new avenues for data exploitation.

“With AI, anyone can create malware. Anyone can find exploits,” Allen said during his XChange August keynote.

For MSPs and solution providers already struggling to keep pace with software vulnerabilities and ransomware threats, Allen said relying solely on patching is no longer enough. "If your solution to this problem is ‘I'm gonna patch all my software,’ that is not going to work,” he said. "These vulnerabilities are being fast found quicker than vendors can patch them."

The rise of agentic AI presents another challenge. Allen demonstrated how AI tools can be manipulated to perform actions that resemble ransomware behavior and warned that AI systems cannot reliably distinguish between legitimate and malicious intent.

"The point is, AI cannot determine intent. And intent is everything," Allen said. “The difference between something that backs up your files and something that exfiltrates data is intent.”

Rather than relying on AI alone to solve AI-driven threats, Allen encouraged partners to focus on zero-trust principles, application allowlisting and behavioral controls. He also emphasized the importance of enforcing a “deny by default, permit by exception” approach to AI access to sensitive resources. He said this strategy is the key to reducing customer risk while still enabling innovation.

Learn how ThreatLocker helps MSPs secure AI adoption and implement zero trust controls.