Skip to content
Back to AI News
The DecoderJuly 15, 2026· 1 min read

OpenAI's Codex now encrypts instructions between AI agents, leaving developers blind to internal delegation

OpenAI's Codex now encrypts instructions between AI agents, making it harder for developers to track task delegation. This change affects larger GPT-5.6 variants Sol and Terra.

What happened

OpenAI's Codex, a coding tool, has started encrypting instructions passed between main and subagents. This means developers can no longer see how tasks are being delegated internally. For larger GPT-5.6 variants Sol and Terra, encryption is now mandatory.

Why it matters

As a business owner, you may rely on Codex for tasks like coding or data processing. With encrypted instructions, you might struggle to understand how tasks are being handled, potentially affecting your workflow and quality control.

The takeaway

You may need to adjust your workflow or find alternative tools to manage task delegation, as OpenAI's Codex changes limit visibility into internal AI agent interactions.

Read the original at The Decoder

Our plain-English take, written from public reporting for operational business owners. Always read the original for full context.

Nayre builds the AI systems behind stories like this.

Chatbots, workflow automation, finance intelligence, and internal knowledge systems. Built for operational teams, shipped in days.