OpenAI's new AI web browser, ChatGPT Atlas, has been found to be vulnerable to clipboard injection attacks. This vulnerability could allow malicious actors to manipulate the user's clipboard, potentially leading to security breaches.

Cybersecurity experts have raised concerns about OpenAI's new browser, ChatGPT Atlas, which may be susceptible to attacks that could compromise user data. The browser's features, including 'browser memories' and 'agent mode,' could potentially be exploited through prompt injection attacks, leading to unauthorized access to sensitive information.