Overview
- Gartner’s advisory recommends blocking AI-native browsers unless organisations complete service-specific risk assessments and implement strict controls with monitoring.
- The report warns that default settings prioritise user experience over security, exposing users to prompt injection, erroneous autonomous actions, phishing-driven credential loss, and costly procurement mistakes.
- Gartner notes that AI sidebars can transmit active page content, open tabs, and browsing history to cloud systems, raising the risk of unintended corporate data exposure.
- Independent studies by SquareX and Cato Networks highlighted exploitable flaws in AI browsers, including malicious workflows and the “HashJack” technique that can weaponise legitimate websites.
- Examples cited include Perplexity’s Comet and OpenAI’s ChatGPT Atlas, as regulators in India study agentic AI’s competition and data risks without yet issuing prescriptive rules.