Pentesting Needed for Our AI Agent

Job ID: 39029490

Budget: $30 – $50 USD

I am in need of a qualified penetration tester to assist in assessing the security of one of our publicly available, computer-use AI agents developed by Anthropic. The primary focus of this task will be to simulate a harmless prompt injection attack on our test AI agent.

Key Details:
- The project involves testing a computer-use AI agent.
- The primary concern is with potential vulnerabilities related to malicious code injection.
- The pentesting should be conducted in our development environment.

Please note that further information will be provided to the successful freelancer we choose to work with. Ideal candidates should have extensive experience with pentesting AI systems, with a particular focus on code injection vulnerabilities.

We estimate this project to only be 2-3 hours worth of work.