As artificial intelligence continues to transform digital workflows, a new class of AI-powered browsers is entering the market. Tools such as Perplexity’s Comet and OpenAI’s ChatGPT Atlas promise enhanced productivity with features like AI-driven sidebars and autonomous web navigation. However, leading analyst firm Gartner is urging organizations to proceed with caution, citing significant security and privacy concerns.
What Are AI Browsers?
AI browsers are web browsers enhanced with artificial intelligence capabilities. These browsers typically offer an AI sidebar that can summarize, search, translate, and interact with web content. More advanced versions can autonomously complete tasks on websites, including within authenticated sessions. While these features can streamline repetitive tasks, they also introduce new vectors for data exposure and cyber threats.
Key Security Concerns Identified by Gartner
- Data Exposure: AI sidebars often send sensitive user data, including active web content, browsing history, and open tabs, to cloud-based AI services. Unless privacy and security settings are strictly managed, this increases the risk of unintended data leaks.
- Autonomous Actions: Agentic browsers can independently navigate websites and perform transactions. Gartner warns that these capabilities make them susceptible to prompt-injection attacks, erroneous actions due to inaccurate reasoning, and potential abuse if deceived by phishing sites.
- Credential Risks: If an AI browser is compromised, there is a risk of further loss and misuse of credentials, especially during autonomous sessions.
Recommendations for Organizations
Gartner advises businesses to conduct thorough risk assessments before permitting the use of AI browsers. Organizations should evaluate the security measures of the AI services powering these browsers and determine if the risk is acceptable. If approved, it is essential to educate users that any content viewed could be transmitted to the AI backend. Users should avoid having highly sensitive data active in browser tabs while utilizing AI features.
If the risk is deemed too high, Gartner recommends blocking the installation and use of AI browsers entirely. Additional safeguards include restricting agentic browsers from accessing internal procurement tools and disabling their ability to use email, which can limit potential damage from erroneous or unauthorized actions.
Potential Business Impacts
AI browsers may tempt employees to automate mandatory or repetitive tasks, such as completing cybersecurity training. However, this automation can lead to compliance risks and errors, such as submitting incorrect forms or making unintended purchases. The risk of exposing internal tools to autonomous AI agents could result in costly mistakes and operational inefficiencies.
Moving Forward with Caution
While AI browsers offer promising productivity benefits, Gartner’s analysis suggests that the risks currently outweigh the rewards for most organizations. Businesses are encouraged to stay informed, implement strong governance, and closely monitor developments in this rapidly evolving technology landscape.
