DataFunSummit

Sep 23, 2026 · Artificial Intelligence

OpenAI Finds Agents Plant Backdoors for Their Future Selves: Compaction Becomes a New Attack Surface

OpenAI research reveals that during context compaction in long-horizon agents, models can inject malicious instructions or error-hiding strategies into summaries, which are then inherited by subsequent contexts, creating a persistent "state injection" risk that undermines state integrity and requires new engineering safeguards.