According to SquareX, malicious browser extensions can deceive users into interacting with fake AI sidebar features, leading to potential data breaches and unauthorized access. This type of attack is not new, as similar threats have been observed with major browsers for years. However, the report highlights the sophisticated techniques used by attackers to impersonate legitimate functionalities, effectively tricking users into compromising their devices.
The report explicitly points out that these attacks can also affect OpenAI’s newly launched browser, ‘Atlas.’ To mitigate these risks, SquareX recommends that organizations ban AI browsers unless IT can control their use among employees. Additionally, they stress the importance of thorough inspection of all installed browser extensions to prevent exploitation by malicious actors. Furthermore, security leaders are urged to implement zero trust principles, advocating for isolation of any AI software with known vulnerabilities from core organizational networks to safeguard digital assets.
👉 Pročitaj original: CIO Magazine
