Safe Recovery of AI Agents in Enterprise Systems: Preventing Duplicate Actions and Preserving Human Corrections
DOI:
https://doi.org/10.54536/ajsts.v5i2.8833Keywords:
AI agents, API contracts, Concurrency control, Fault recovery, Workflow reliabilityAbstract
Reliable use of AI agents in enterprise systems depends on what happens after an action fails or its outcome becomes uncertain. Retrying can duplicate an order, while reversing a workflow can invalidate a person’s subsequent correction. The central challenge is deciding whether recovery is justified by the available evidence and still satisfies the business request. We investigate this challenge in a reproducible testbed connecting order, inventory and billing services with different API guarantees. Controlled failures and concurrent edits reveal two evaluation gaps: protecting individual records does not ensure consistency across related services, and preserving an edit does not ensure that the final workflow fulfils its instruction. A separate diagnostic identified nine apparent completions that left human instructions unsatisfied. A conventional recovery implementation with equivalent safeguards matched the evaluated controller in all 270 paired runs, locating the observed benefit in shared engineering protections rather than a distinct recovery algorithm. On the prospective test configurations, both completed 108 of 180 runs without scored violations and left 72 for manual resolution. These findings provide a practical basis for evaluating agent integrations: verify uncertain effects, check related records, and distinguish completed business goals from unresolved handoffs. The contribution is a reproducible evaluation of these recovery requirements within a synthetic enterprise workflow, with operational effectiveness remaining to be established.
Downloads
References
Barres, V., Dong, H., Ray, S., Si, X., & Narasimhan, K. (2025). τ²-Bench: Evaluating conversational agents in a dual-control environment. arXiv. https://arxiv.org/abs/2506.07982v1
Birrell, A. D., & Nelson, B. J. (1984). Implementing remote procedure calls. ACM Transactions on Computer Systems, 2(1), 39–59. https://doi.org/10.1145/2080.357392
Chang, E. Y. (2026). TRW: TRACE-RealWorld—An auditable consistency contract for world models as materialized views. arXiv. https://arxiv.org/abs/2607.21910v2
Chang, E. Y., & Geng, L. (2025). SagaLLM: Context management, validation, and transaction guarantees for multi-agent LLM planning. arXiv. https://arxiv.org/abs/2503.11951v3
Chen, Z., Liu, H., Xu, D., Dong, D., Li, J., Pu, B., & Zhai, J. (2026). Cordon: Semantic transactions for tool-using LLM agents. arXiv. https://doi.org/10.48550/arXiv.2606.17573
Featonby, M. (n.d.). Making retries safe with idempotent APIs. Amazon Builders’ Library. https://aws.amazon.com/builders-library/making-retries-safe-with-idempotent-APIs/
Fielding, R., Nottingham, M., & Reschke, J. (2022). HTTP semantics (RFC 9110). Internet Engineering Task Force. https://doi.org/10.17487/RFC9110
Garcia-Molina, H., & Salem, K. (1987). Sagas. In Proceedings of the 1987 ACM SIGMOD international conference on management of data (pp. 249–259). ACM Press. https://doi.org/10.1145/38713.38742
Gray, J., & Lamport, L. (2006). Consensus on transaction commit. ACM Transactions on Database Systems, 31(1), 133–160. https://doi.org/10.1145/1132863.1132867
Gunawi, H. S., Do, T., Joshi, P., Alvaro, P., Hellerstein, J. M., Arpaci-Dusseau, A. C., Arpaci-Dusseau, R. H., Sen, K., & Borthakur, D. (2011). FATE and DESTINI: A framework for cloud recovery testing. In 8th USENIX Symposium on Networked Systems Design and Implementation (NSDI 11). USENIX Association. https://www.usenix.org/conference/nsdi11/fate-and-destini-framework-cloud-recovery-testing
Herlihy, M. P., & Wing, J. M. (1990). Linearizability: A correctness condition for concurrent objects. ACM Transactions on Programming Languages and Systems, 12(3), 463–492. https://www.cs.cmu.edu/~wing/publications/HerlihyWing90.pdf
Korth, H. F., Levy, E., & Silberschatz, A. (1990). A formal approach to recovery by compensating transactions. In D. McLeod, R. Sacks-Davis, & H. Schek (Eds.), Proceedings of the 16th international conference on very large data bases (pp. 95–106). Morgan Kaufmann. https://www.vldb.org/conf/1990/P095.PDF
Kung, H. T., & Robinson, J. T. (1981). On optimistic methods for concurrency control. ACM Transactions on Database Systems, 6(2), 213–226. https://doi.org/10.1145/319566.319567
Liu, X., Yu, H., Zhang, H., Xu, Y., Lei, X., Lai, H., Gu, Y., Ding, H., Men, K., Yang, K., Zhang, S., Deng, X., Zeng, A., Du, Z., Zhang, C., Shen, S., Zhang, T., Su, Y., Sun, H., … Tang, J. (2024). AgentBench: Evaluating LLMs as agents. In The Twelfth International Conference on Learning Representations. https://proceedings.iclr.cc/paper_files/paper/2024/hash/e9df36b21ff4ee211a8b71ee8b7e9f57-Abstract-Conference.html
Lyu, H., Zhang, D., Wu, M., Wei, X., & Chen, H. (2026). CoAgent: Concurrency control for multi-agent systems. arXiv. https://arxiv.org/abs/2606.15376v1
Mansoor, I. K., Phadke, A., & Rana, P. (2026). Verified tool calls improve LLM agent reliability under non-atomic failures. arXiv. https://arxiv.org/abs/2608.02645v1
Microsoft. (n.d.). Compensating transaction pattern. Azure Architecture Center. https://learn.microsoft.com/en-us/azure/architecture/patterns/compensating-transaction
Microsoft. (2024). Perform conditional operations using the Web API. Microsoft Dataverse documentation. https://learn.microsoft.com/en-us/power-apps/developer/data-platform/webapi/perform-conditional-operations-using-web-api
Mohammadi, B., Potamitis, N., Klein, L., Arora, A., & Bindschaedler, L. (2026). Atomix: Timely, transactional tool use for reliable agentic workflows. arXiv. https://arxiv.org/abs/2602.14849v2
National Institute of Standards and Technology. (2026, February 17). Announcing the “AI Agent Standards Initiative” for interoperable and secure innovation. https://www.nist.gov/news-events/news/2026/02/announcing-ai-agent-standards-initiative-interoperable-and-secure
Perera, S., Hapuarachchi, K., Leymann, F., & Khalaf, R. (2026). Robust agent compensation (RAC): Teaching AI agents to compensate. In Proceedings of the ACM Conference on AI and Agentic Systems (pp. 253–262). Association for Computing Machinery. https://doi.org/10.1145/3786335.3813141
Stripe. (n.d.). Idempotent requests. https://docs.stripe.com/api/idempotent_requests
Sun, Z., Wang, X., & Li, G. (2026). Agentic transaction: Towards ACID-compliant agent systems. arXiv. https://arxiv.org/abs/2608.13900v1
Tong, Y., Dai, L., & Guo, S. (2026). AID-Guard: Stateful authorization for delegated agent effects. arXiv. https://arxiv.org/abs/2608.21159v1
Wang, A., Hager, S., Asija, A., Khashabi, D., & Andrews, N. (2025). Hell or high water: Evaluating agentic recovery from external failures. In Second Conference on Language Modeling. https://doi.org/10.48550/arXiv.2508.11027
Wu, J., Chen, X., & Shi, R. (2026). A proactive fault tolerance approach for Saga transactions in microservices using a health prediction model. Journal of Universal Computer Science, 32(8), 1150–1172. https://doi.org/10.3897/jucs.175609
Yao, S., Shinn, N., Razavi, P., & Narasimhan, K. (2024). τ-bench: A benchmark for tool-agent-user interaction in real-world domains. arXiv. https://arxiv.org/abs/2406.12045v1
Yao, S., Zhao, J., Yu, D., Du, N., Shafran, I., Narasimhan, K., & Cao, Y. (2023). ReAct: Synergizing reasoning and acting in language models. In The Eleventh International Conference on Learning Representations. https://doi.org/10.48550/arXiv.2210.03629
Yuvin Labs. (n.d.). ConsequenceBench [Computer software]. GitHub. https://github.com/yuvin-labs/consequencebench
Downloads
Published
Issue
Section
License
Copyright (c) 2026 Medhavee Upadhyaya, Gaurav Chaulagain
This work is licensed under a Creative Commons Attribution 4.0 International License.
