1. New AI-powered browsers like OpenAI's ChatGPT Atlas and Perplexity's Comet aim to replace Google Chrome as the primary internet gateway, using AI agents for tasks on behalf of users.

2. However, these AI browser agents pose a greater risk to user privacy compared to traditional browsers due to potential "prompt injection attacks", where bad actors can trick AI agents into executing malicious commands and exposing sensitive data.

3. To mitigate these risks, users should consider limiting the access they give AI browser agents, use unique passwords and multi-factor authentication for these accounts, and avoid giving them control over sensitive information until security around these tools improves.