Codex Encryption Update
Executive Summary
- Codex is encrypting prompts for inter-agent communications.
- Inference is still done in plaintext after decryption on the server side.
- This change aims to hide communications from the end-user.
The Buzz Score
The Internet’s Verdict: 70% Hyped, 30% Skeptical
Understanding the Update
Codex’s new approach has sparked discussion.
HN Title is very misleading, it makes it sound like inference is being done directly on ciphertext, which would require homomorphic encryption well advanced of what is known.
The encryption happens when an agent calls into another agent without exposing the plaintext to the end-user.
This title seems extremely easy to misinterpret. If I understand correctly: Codex is now encrypting inter-agent communications and hiding those communications from the user.
Implications and Questions
Some wonder if this change will affect using GPT subscriptions in alternative harnesses.
The location of the encryption is also a point of interest.
Could someone explain to me where exactly the encryption is happening? I assumed that the main agent makes calls to sub-agents locally.
Focus Keyword: Codex Encryption
