KnowBe4, the world-renowned cybersecurity platform that comprehensively addresses human and AI agent risk management, announced breakthrough innovations that train both people and AI agents to defend against evolving cyber threats. According to Gartner, forty percent of enterprise applications will be integrated with task-specific AI agents by the end of 2026, up from less than 5% today. These AI agents are the new targets for theft, adversarial manipulation and misuse, which emphasizes the importance of employee AI literacy.
For years, the human element has been involved in over 60% of breaches, including social engineering as one of the top attack vectors, confirmed in reports such as the 2025 Verizon Data Breach Investigations Report. Given the prevalence of the human element in breaches, smarter defenses that address the dynamic between humans and AI agents is integral to build a solid cybersecurity defense strategy.
“We are delivering adaptive, personalized, AI-driven defenses today to safeguard the next generation global workforce,” said Bryan Palma, CEO, KnowBe4. “We do both, guard humans against AI-driven attacks and shield AI systems from exploitation. KnowBe4 is uniquely positioned to protect enterprise AI agents by securing the essential human-agent interaction layer. As AI becomes both a powerful tool and a potential target, safeguarding this interaction point is crucial.”
Also Read: Coupa Announces Acquisition of AI-Powered Scoutbee to Drive Supplier Intelligence and Discovery
Cyber threats are growing more sophisticated through AI, but AI itself is also becoming a tool and a high-value target. While most solutions focus solely on defending at the gateway layer, KnowBe4’s HRM+ goes further, delivering true defense-in-depth. Built on 15 years of user behavior and threat intelligence data, the platform ensures that protection extends beyond the edge, securing every layer of interaction, from the productivity environment to the individual user and the AI agents themselves. This layered approach provides resilience that no other platform currently matches and includes:
- Agent-Safe Behavior Training: Just as employees learned to spot a malicious link, they must now learn how to safely interact with and oversee AI agents.
- Prompt Injection & Manipulation Defense: Simulated attacks train global workforces to identify and resist adversarial inputs designed to hijack enterprise AI agents.
- Risk Scoring for Agent Interactions: Extending the industry-leading Risk Score to measure susceptibility to agent misuse provides comprehensive risk quantification.
Source: BusinessWire