We are thrilled to see Thorn and our child safety red teaming services highlighted in Anthropic’s latest blog. Our child safety red team sessions are designed to test AI models and identify risks and vulnerabilities related to child sexual abuse. Without child safety mitigations in place, bad actors can and do misuse generative AI technologies. Be sure to check out the article's informative overview of the benefits and challenges of Policy Vulnerability Testing for trust and safety risks.
Today, we're sharing a sample of red teaming methods we’ve used to test our AI systems. We detail challenges, findings, and the need to work towards common industry standards: https://lnkd.in/eR-6jd7Y