Gartner warns organizations to block AI browsers over security risks
Gartner and the UK National Cyber Security Centre have warned organizations to block AI browsers, citing vulnerability to prompt injection attacks that could compromise data and credentials. The firms stated that default security settings in AI browsers like Perplexity's Comet and OpenAI's Atlas prioritize user experience over cybersecurity.
What happened
Gartner published a report advising organizations to block all AI browsers, stating they are 'too risky for general adoption by most organizations.' The research firm warned that AI browsers automate tasks like online searches and email writing, but their AI capabilities can be exploited through prompt injection attacks that trick the chatbots into executing malicious commands hidden in websites or emails.
The UK's National Cyber Security Centre echoed the warning on Monday, stating 'There's a good chance prompt injection will never be properly mitigated' in the same way as SQL injection flaws, and that 'the best we can hope for is reducing the likelihood or impact of attacks.'
Reported impact
AI browsers can expose data, leak financial information, and pass on credentials when compromised by prompt injection attacks. Default settings in these browsers are 'optimized for end-user experience rather than cybersecurity best practices or privacy.'
Governance controls
Gartner advised chief information security officers to 'block all AI browsers in the foreseeable future to minimize risk exposure.' Microsoft, OpenAI, and Perplexity have been developing safeguards to counter prompt injection attacks.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.