Journal article · 2026 · Open access

Securing LLM-based agents against cyberattacks: a comprehensive survey on attack techniques and defense strategies

Nyashadzashe Tamuka, Topside Ehleketani Mathonsi, Thomas Otieno Olwal, Solly P. Maswikaneng, Tonderai Muchenje, Tshimangadzo Mavin Tshilongamulenzhe

Journal of Computer Virology and Hacking Techniques, 22(1) · Published 7 May 2026 · DOI 10.1007/s11416-026-00622-3

Open PDFView at the publisherCite this paper

In brief

“Securing LLM-based agents against cyberattacks: a comprehensive survey on attack techniques and defense strategies” is a 2026 journal article by Nyashadzashe Tamuka and 5 co-authors, published in Journal of Computer Virology and Hacking Techniques. OpenAlex counts 14 citations as of 19 November 2026, which puts it in the top 2% of works published in 2026 by citations. It is open access under a CC BY licence: the full text is free to read at Journal of Computer Virology and Hacking Techniques.

Citations
14
Most cited in
2026 (14)
Field-weighted impact
15.14
References
60
Authors
6

Abstract

Large Language Model (LLM)-based agents integrate various models, including planning loops, memory, tool use, and multi-agent systems, enabling autonomous decision-making through natural-language interfaces. This autonomy also expands the cyberattack surface from model-only failures to agent compromise, where untrusted text can exfiltrate data or trigger malicious actions. This survey presents a taxonomy-driven synthesis of security threats targeting LLM-based agents and a structured review of various studies, summarizing each work’s aims, methods, results, and limitations and mapping them to the proposed taxonomy. Attacks, including context and prompt manipulation, jailbreak and cognition attacks, privacy attacks, agentic action induction and availability, and reconnaissance threats, were categorized. The widely adopted defenses are classified into prompt/input sanitization, adversarial training and alignment tuning, architectural safeguards, monitoring and guardrails, and watermarking/provenance mechanisms for source verification. The persistent gaps were a weak generalization of defense under attacker adaptation and a limited evaluation of agent-tool interaction risk. Since real deployments increasingly rely on agents that can remember and act, security must shift from single-model filtering to system-level, benchmarked, and continuously tested controls that measure resilience under adaptive attackers. Open challenges and research directions include robust detection of covert prompt attacks, continuous red-teaming, measurable security utility trade-offs, provenance-aware agent pipelines, and standardized protocols for evaluating secure LLM agents in real deployments.

Cite this paper

Formatted from the catalogue record. Check names, capitals and page numbers against the paper before you submit.

APA 7

Tamuka, N., Mathonsi, T. E., Olwal, T. O., Maswikaneng, S. P., Muchenje, T., & Tshilongamulenzhe, T. M. (2026). Securing LLM-based agents against cyberattacks: a comprehensive survey on attack techniques and defense strategies. Journal of Computer Virology and Hacking Techniques, 22(1). https://doi.org/10.1007/s11416-026-00622-3

MLA 9

Tamuka, Nyashadzashe, et al. “Securing LLM-based agents against cyberattacks: a comprehensive survey on attack techniques and defense strategies.” Journal of Computer Virology and Hacking Techniques, vol. 22, no. 1, 2026. https://doi.org/10.1007/s11416-026-00622-3.

Chicago

Tamuka, Nyashadzashe, Topside Ehleketani Mathonsi, Thomas Otieno Olwal, Solly P. Maswikaneng, Tonderai Muchenje, and Tshimangadzo Mavin Tshilongamulenzhe. 2026. “Securing LLM-based agents against cyberattacks: a comprehensive survey on attack techniques and defense strategies.” Journal of Computer Virology and Hacking Techniques 22 (1). https://doi.org/10.1007/s11416-026-00622-3.

Harvard

Tamuka, N. et al. (2026) ‘Securing LLM-based agents against cyberattacks: a comprehensive survey on attack techniques and defense strategies’, Journal of Computer Virology and Hacking Techniques, 22(1). doi:10.1007/s11416-026-00622-3.

IEEE

N. Tamuka, T. E. Mathonsi, T. O. Olwal, S. P. Maswikaneng, T. Muchenje, and T. M. Tshilongamulenzhe, “Securing LLM-based agents against cyberattacks: a comprehensive survey on attack techniques and defense strategies,” Journal of Computer Virology and Hacking Techniques, vol. 22, no. 1, 2026, doi: 10.1007/s11416-026-00622-3.

BibTeX

@article{tamuka2026securing,
  title     = {{Securing LLM-based agents against cyberattacks: a comprehensive survey on attack techniques and defense strategies}},
  author    = {Tamuka, Nyashadzashe and Mathonsi, Topside Ehleketani and Olwal, Thomas Otieno and Maswikaneng, Solly P. and Muchenje, Tonderai and Tshilongamulenzhe, Tshimangadzo Mavin},
  journal   = {Journal of Computer Virology and Hacking Techniques},
  year      = {2026},
  volume    = {22},
  number    = {1},
  doi       = {10.1007/s11416-026-00622-3},
  url       = {https://doi.org/10.1007/s11416-026-00622-3}
}

Download for a reference manager: BibTeX (.bib) · RIS for Zotero, Mendeley and EndNote (.ris)

Topic and keywords

Physical Sciences › Computer Science › Artificial Intelligence › Adversarial Robustness in Machine Learning

  • LLM agents
  • agent security
  • attack taxonomy
  • prompt manipulation
  • jailbreak attacks
  • privacy attacks
  • availability attacks
  • reconnaissance attacks

Work with this paper

Most-cited papers in Computer Science

All highly cited Computer Science papers

Questions about this paper

How do I cite this paper in APA style?

Tamuka, N., Mathonsi, T. E., Olwal, T. O., Maswikaneng, S. P., Muchenje, T., & Tshilongamulenzhe, T. M. (2026). Securing LLM-based agents against cyberattacks: a comprehensive survey on attack techniques and defense strategies. Journal of Computer Virology and Hacking Techniques, 22(1). https://doi.org/10.1007/s11416-026-00622-3

How many times has this paper been cited?

14 times, according to OpenAlex on 19 November 2026. Google Scholar and Scopus count different sets of sources, so their numbers differ.

Is this paper free to read?

Yes. It is open access at Journal of Computer Virology and Hacking Techniques: sign in and open the PDF from this page.

Who wrote this paper?

Nyashadzashe Tamuka, Topside Ehleketani Mathonsi, Thomas Otieno Olwal, Solly P. Maswikaneng, Tonderai Muchenje, Tshimangadzo Mavin Tshilongamulenzhe.

What is the DOI of this paper?

10.1007/s11416-026-00622-3, permanently at https://doi.org/10.1007/s11416-026-00622-3. Sign in and open it from this page.

Bibliographic data from OpenAlex (CC0), last checked 19 November 2026. ResearchGuru does not host this paper; once you are signed in, the links open it at the publisher or the repository that does.