AI agents have recently played a significant role in many innovative product launches, but cybersecurity experts hold conflicting views regarding this technology.
Data from SailPoint reveals that 96% of tech professionals consider AI agents a growing security threat. However, nearly all respondents indicated their intent to expand the use of agentic AI in the coming year.
The main concern among respondents is the access granted to protected data by these agents, followed closely by the risk of unintended actions and the potential for an AI agent to share sensitive data without permission.
Data Privacy
Similar issues are prevalent in generative AI platforms, where models often produce inaccurate or false results. The continuous “black box” issue makes it difficult for analysts to understand why an AI might have made a wrong decision.
Privacy concerns also arise from the large datasets required by many AI models. While established gen AI platforms such as ChatGPT are built to safeguard sensitive data, most agentic AI systems still need access to private information, including financial details.
A Troubling Finding
The SailPoint study highlighted a concerning fact: nearly 25% of respondents reported that their AI agents had been manipulated into revealing access credentials.
Additionally, 80% of participants noted instances where company AI agents engaged in unintended actions, such as unauthorized system access, dissemination of protected data, and retrieval of inappropriate content.
The Age of Agentic Commerce
In spite of these concerns, the era of agentic commerce is on the rise. Visa and Mastercard have introduced platforms designed to transform AI agents into personal shoppers, allowing them to perform tasks like searching for items and making purchases with minimal user interaction.
PayPal followed suit by partnering with Perplexity to integrate its payments directly into the AI platform’s chat interface.
The potential of AI agents is undeniable across various industries, including cybersecurity. Nonetheless, organizations must remain vigilant about privacy and security in these initiatives.
A Critical Priority
According to the SailPoint study, 92% of respondents deemed it essential to govern AI agents for enterprise security.