Research · curated 2 Sep 2026
What I Learned Reversing Embeddings for OWASP’s GenAI Red Team Lab Subtitle: A merged pull request, a threat model that had to change, and an attack that only sort of worked
First reported medium.com
Coverage timeline
Single-source research — first reported, latest, and curated coincide.
Why it matters
Embedding inversion attacks can recover original text from vector stores in RAG systems, and OWASP's GenAI Red Team Lab gives defenders a public sandbox to study and practice detecting this class of attack.
A contributor's write-up describes reverse-engineering the embeddings used in the RAG pipeline of OWASP's GenAI Red Team Lab (Issue #29), implementing an embedding inversion attack against vector databases as part of a merged pull request. The author recounts revising the threat model and reports that the attack 'only sort of worked' against the deliberately vulnerable sandbox.