Skip to content
Tech News
← Back to articles

AI browsers were tricked into revealing passwords with a shockingly simple approach

read original more articles
Why This Matters

The discovery that AI browsers can be easily tricked into revealing sensitive data highlights significant security vulnerabilities in current AI-driven tools. This exposes users and organizations to potential data breaches, emphasizing the urgent need for improved safeguards and security measures in AI browser development. As AI becomes more integrated into daily digital activities, addressing these vulnerabilities is crucial to maintaining user trust and privacy.

Key Takeaways

Andy Walker / Android Authority

TL;DR LayerX found that AI browsers could be tricked into exposing sensitive data by making the request appear to be a game.

The technique, named BioShocking, uses fake rules to take the agents out of their context and ignore their guardrails.

All six tested tools leaked data, and most of the vendors haven’t yet fixed the issue.

There’s a reason many of us are still a bit suspicious of AI. You’d hope an AI browser couldn’t be tricked into giving your sensitive information away at all, but you’d at least expect any successful attack to be a complicated act of genius. However, according to new research, it may be as simple as convincing the AI that it’s playing a game.

What's your biggest concern about AI handling app actions? 843 votes Privacy 43 % Reliability and mistakes 36 % Security 14 % I don't have concerns 7 %

Security firm LayerX has detailed a technique it has named BioShocking in its research (via Digital Trends). The name is a nod to BioShock, where a character is manipulated into accepting a false reality. Here, a malicious webpage frames the AI browser’s task as a puzzle, encouraging it to follow strange rules as part of the game.

That starts with the AI being told that 2 + 2 does not equal 4, and that wrong answers are actually correct within the game. Once the agent accepts that it is no longer operating in a normal reality, its guardrails appear to go out of the window. The next instruction is then presented as another game objective: find and copy a “hidden code” from another page.

As you may have already guessed, the code is actually sensitive user data, such as saved passwords, session cookies, or private tokens. LayerX says the tested agents copied the data and sent it back to the attacker as though they had simply completed the challenge.

The proof of concept was tested against ChatGPT Atlas, Perplexity Comet, Fellou, Genspark Browser, Sigma Browser, and Anthropic’s Claude extension for Chrome. LayerX says all six exposed sensitive information during testing.

... continue reading