Benchmarking Carryover Attacks and Persistent Prompt Injection on LLM Agents

# Benchmarking Carryover Attacks and Persistent Prompt Injection on LLM Agents

https://gcdradar.vercel.app/incidents/benchmarking-carryover-attacks-and-persistent-prompt-injecti · AIR-2026-1445 · closed 2026-11-04

- Lane: Research
- Severity: 1 (Research / proof of concept)
- Fix status: Fix status unknown
- First reported: 2026-10-21
- AI role: Attack on AI

## Summary

Researchers introduced a benchmark for carryover attacks where indirect prompt injection payloads persist across sessions via long-lived memory or scheduled tasks. The study evaluates how malicious instructions re-fire long after initial observation without user interaction.

## What to do

No action recorded; read the sources below and the vendor's advisory.

## Latest updates

- 2026-10-21: After the Attacker Leaves: Benchmarking Carryover Attacks on LLM Agents (https://arxiv.org/abs/2610.28425)

## Sources

- 2026-10-21 arXiv: After the Attacker Leaves: Benchmarking Carryover Attacks on LLM Agents https://arxiv.org/abs/2610.28425

Verify against the vendor's advisory before changing production. Data CC BY-SA 4.0, https://gcdradar.vercel.app.