Gartner Warns: AI Browsers Pose Major Enterprises Security Risks

Gartner has issued a rare and explicit warning to enterprises: avoid deploying AI browsers such as Comet, Atlas, or other agentic browsing tools. The advisory, released in a new report by Gartner analysts, states that AI-native browsers present significant cybersecurity risks, especially for organizations handling sensitive data. 

According to the report, AI browsers fundamentally alter long-held security assumptions by continuously analyzing open tabs, scraping visible data, and transmitting contextual information to external AI back-ends. This creates new exposure pathways for confidential documents, internal dashboards, credentials, and regulated datasets. Gartner concluded that CISOs should block all AI browsers in the foreseeable future to limit risk. 

Security researchers echoed these concerns. Experts note that AI browsers can unintentionally collect data across tabs, interpret hidden instructions, and even take autonomous actions. For example, clicking links, filling out forms, or sending information, without the user realizing it.
Additional risks highlighted include: 

• Cross-tab visibility: AI models gain full insight into all open browser content. 

• Autonomous actions: AI assistants can execute harmful instructions embedded in webpages. 

• Fingerprinting risks: Attackers can easily detect AI browsers at scale for targeted attacks. 

• Shadow IT proliferation: Employees may install AI browsers at home and bring them into workplace environments via BYOD or synced accounts. 

Experts warn AI browsers are evolving faster than security guardrails can keep pace. Proprietary AI back-ends remain “black boxes,” making audits nearly impossible. Although training employees on safe usage can help, analysts argue that awareness alone cannot mitigate systemic risks. 

Gartner recommends organizations carefully evaluate any AI browser’s data-handling practices, impose strict DLP and access controls, and prepare for emerging threats as agent-driven browsing accelerates. For now, however, the firm’s stance is clear: AI browsers introduce more risk than benefit in enterprise settings. 

 

Source: 

https://www.technewsworld.com/story/gartner-recommends-avoiding-ai-browsers-for-now-180050.html