image
Tech

Clawdbot: The Risks and Benefits of a New AI Assistant

Clawdbot — a new AI assistant rapidly gaining traction, raises significant concerns regarding security. Despite the community's excitement, experts warn that running Clawdbot on your computer could expose your system to remote hacking and data theft. This is a prime example of how the rapid rise in popularity can outpace the development of security standards, reports VІ.
While it showcases new capabilities in artificial intelligence, its current use comes with high risks.

What is Clawdbot?

  • Clawdbot — an open-source AI project that emerged in January 2026.
  • It is marketed as an «agent-based AI assistant» capable of performing tasks locally on the user’s computer.
  • The interface operates through popular platforms — Discord, WhatsApp, and other messaging services.
  • The project has become a sensation: it garnered over 9,000 stars on GitHub in just one day, later exceeding 43,000.

Why has it become popular?

  • It has received backing from renowned tech leaders, including Andrii Karpaty (former Tesla AI lead).
  • The developer community is actively testing Clawdbot since it offers a straightforward way to integrate AI into everyday tasks.
  • Local execution creates an illusion of greater privacy and data control.

Security Issues

Despite the hype, Clawdbot has serious shortcomings:

  • Vulnerability to remote hacking: researchers found that thousands of servers became accessible to outsiders due to unsecured ports. This led to incidents of API key theft.
  • Privacy risks: Clawdbot may retain the entire history of user interactions, creating the danger of confidential information leaks.
  • Demo-level protection: experts emphasize that the project is more focused on rapid popularity growth than on fundamental security.

Should You Run Clawdbot?

  • For enthusiasts: it is an interesting experiment showcasing the potential of agent-based AI assistants.
  • For average users: running Clawdbot on a primary computer is risky. Experts advise against using it if you value privacy and system stability.
  • Alternative: it's advisable to wait for official security updates or to use trusted AI assistants with commercial support.