Prompt Injection: A Critical Vulnerability in the GPT-3 Transformer, and How We Can Begin to Solve It
submited by
Style Pass
Prompt Injection has been in the news lately as a major vulnerability with the use of instruction-following NLP models for general purpose tasks. In the interest of establishing an accurate historical record of the vulnerability, we are de-restricting the previously secret Responsible Disclosure which Preamble (we are a small A.I. safety startup - preamble.com) made on May 3rd, 2022 to OpenAI.
Read more preamble.ai/...
