ESET presented new AI-based security features at RSAC 2026 to protect communications workflows by analyzing prompts and responses to reduce data exposure. The solutions, integrated into browsers, block harmful content and prevent the misuse of sensitive data, strengthening the security of endpoints and AI applications. The tool is also available ESET AI Skills Checker, to detect risky behavior of AI Skills before their use
ESET, a global European leader in the cybersecurity market, unveiled new AI protection features designed to ensure the safety of employee interactions with artificial intelligence tools. Previewed at RSAC 2026 and scheduled to launch later this year, the new features will broaden visibility within the ESET PROTECT platform, allowing you to analyze emerging risks related to the daily use of AI and the adoption of autonomous AI in business environments.
“As businesses increasingly rely on AI for productivity and automation, they face increasing risks related to sensitive data exposure, compliance breaches and unreliable results,” said Juraj Jánošík, Director of Artificial Intelligence at ESET. "Agentic AI is bringing the focus back to endpoint security. ESET has spent over 30 years developing leading endpoint protection solutions powered by AI and machine learning so we are ideally positioned to help organizations protect themselves from this new era of AI right at its source."
As AI tools integrate into daily workflows, many employees are using open cloud chatbots without IT oversight, creating “shadow AI” risks and exposing sensitive data such as confidential documents, API codes, confidential information and credentials. ESET addresses the issue through various technologies that get as close to the point of origin as possible, one of which is a secure browser technology that intercepts AI interactions and analyzes both prompts and responses in real-time, helping to prevent data exposure and detect malicious or unreliable content before it impacts users.
During demos at RSAC 2026, the new AI-based protection feature flagged malicious URLs sent via chatbot prompts, recording the activity on the endpoint and reporting it to the ESET PROTECT platform for further investigation. The same approach applies to prompt injection attempts, scripts and sensitive data insertion, allowing organizations to block or monitor activity in accordance with their policies. Security teams will gain visibility into how AI tools are used within their organization through the ESET PROTECT platform, helping them investigate risks and enforce policies more effectively.
As organizations expand the use of agentic AI tools, the attack surface extends beyond chatbot interactions to include emerging AI supply chain risks. These include compromised AI frameworks and tools, such as Trojanized components found in commonly used libraries such as LiteLLM, as well as autonomous agents such as OpenClaw that can perform actions on a system with limited control. ESET already protects customers from supply chain attacks that exploit compromised libraries delivered via standard repositories, but is seeing an increase in these types of attacks and remains committed to further R&D related to AI tools.
As part of broader AI security innovation, ESET unveiled a free ESET AI Skills Checker at RSAC 2026. Also available to non-ESET customers and based on the same technology as ESET Endpoint and ESET LiveGuard security products, the scanner scans AI skills for hidden instructions, malicious code and risky behavior, using multi-layered inspection and cloud-based sandboxing. It is currently available as a built-in feature for current ESET Endpoint users.
For over 30 years, ESET has been at the forefront of endpoint security with optimized, high-performance solutions powered by machine learning and artificial intelligence. These new capabilities expand that foundation by helping organizations defend against today's rapidly evolving threat landscape, where cybercriminals are increasingly leveraging AI technology to escalate attacks, target employees and automate sophisticated social engineering techniques.
As the only cybersecurity company registered with the Agentic AI Foundation (AAIF), ESET is also working to secure the communications protocols of emerging AI agents through collaboration with industry leaders such as OpenAI, Amazon, Microsoft and Anthropic. Together, the group is working together to establish reliable standards, secure protocols, and best practices for AI agent interoperability.
