Tencent's Team Memory shares AI agent memory across a team — with no governance yet for when it's wrong

Tencent expanded its open-source Agent Memory project with a beta feature called Team Memory, which lets multiple AI agents pull from a shared context hub rather than isolated session memory. The system includes access controls and reusable assets such as chat history and persona context, after Tencent said its persona layer improved long-session accuracy from 48% to 76%.

Shared memory could make enterprise agent workflows more coherent, especially as companies try to reduce hallucinations caused by missing or inconsistent context. But it also creates a new risk: if wrong context enters the shared layer, errors can propagate across an entire agent team, making governance and correction mechanisms critical.