PUB
  • Home
  • Sources
  • Search
  • About

Cybersecurity Responsibilities in Enterprise Generative AI Adoption: A Qualitative Documentary Analysis of Delegated Authority and Organizational Oversight

Xiaohang Li, Chengzhi Cai, Haizhou Wang
Academic Journal of Management Science and Engineering, (2026), Vol.1, No.7, pp.1-17
Published: October 12, 2026
DOI: 10.62051/ajmse.v1n7.01
PDF: Download Full Text PDF
Abstract

Enterprise generative artificial intelligence assistants can retrieve organizational information, recommend decisions, and initiate actions through connected tools. These capabilities make cybersecurity a question of how authority and responsibility are distributed across systems and organizations. This study examines how public guidance translates security responsibilities into controls over enterprise AI assistants. A purposive corpus of 12 documentary units from public bodies, a security community, and technology providers was retrieved on 28 September 2026. Focused qualitative document analysis combined six initial coding categories with cross-document comparison, producing a register of 27 evidence records. Four interpretive themes were developed: responsibility is allocated at different organizational and technical levels; permission inheritance requires contextual judgment about appropriate access; human oversight depends on enforceable intervention points; and incident accountability requires continuity across organizational and temporal boundaries. Counterevidence matters: the corpus contains explicit oversight, authorization, and incident-management provisions, so it does not support a general claim that AI security guidance lacks responsibility requirements. The analysis instead identifies an implementation translation problem between broad obligations, application-level controls, and locally assigned decision rights. A proposed governance framework links retrieval, recommendation, and action capabilities to permission boundaries, intervention conditions, accountable owners, and retained decision evidence. The framework is an analytical synthesis, not an empirically validated maturity model. The study contributes a focused account of responsibility at transitions between AI capabilities and offers an auditable structure for enterprise deployment decisions. Its conclusions concern the selected documents and do not establish organizational compliance, control effectiveness, or industry-wide prevalence.

Keywords: Generative Artificial Intelligence, Cybersecurity Governance, Enterprise AI Assistants, Delegated Authority, Human Oversight, Qualitative Document Analysis
APA Citation: Xiaohang Li, Chengzhi Cai, Haizhou Wang (2026). Cybersecurity Responsibilities in Enterprise Generative AI Adoption: A Qualitative Documentary Analysis of Delegated Authority and Organizational Oversight. Academic Journal of Management Science and Engineering, 1(7), 1-17. https://doi.org/10.62051/ajmse.v1n7.01

References

  1. Amazon Web Services. (n.d.-a). Security, guardrails, and observability in Amazon Bedrock. Amazon Bedrock User Guide. Retrieved September 28, 2026, from https://docs.aws.amazon.com/bedrock/latest/userguide/security.html.
  2. Amazon Web Services. (n.d.-b). Incident response in Amazon Bedrock. Amazon Bedrock User Guide. Retrieved September 28, 2026, from https://docs.aws.amazon.com/bedrock/latest/userguide/security-incident-response.html.
  3. Bowen, G. A. (2009). Document analysis as a qualitative research method. Qualitative Research Journal, 9(2), 27–40. https://doi.org/10.3316/QRJ0902027.
  4. Buçinca, Z., Malaya, M. B., & Gajos, K. Z. (2021). To trust or to think: Cognitive forcing functions can reduce overreliance on AI in AI assisted decision making. Proceedings of the ACM on Human Computer Interaction, 5(CSCW1), 188. https://doi.org/10.1145/3449287.
  5. Debenedetti, E., Zhang, J., Balunović, M., Beurer Kellner, L., Fischer, M., & Tramèr, F. (2024). AgentDojo: A dynamic environment to evaluate prompt injection attacks and defenses for LLM agents. Advances in Neural Information Processing Systems, 37. https://proceedings.nips.cc/paper_files/paper/2024/hash/97091a5177d8dc64b1da8bf3e1f6fb54 Abstract Datasets_and_Benchmarks_Track.html.
  6. Debenedetti, E., Shumailov, I., Fan, T., Hayes, J., Carlini, N., Fabian, D., Kern, C., Shi, C., Terzis, A., & Tramèr, F. (2025). Defeating prompt injections by design. arXiv preprint arXiv:2503.18813. https://doi.org/10.48550/arXiv.2503.18813.
  7. Department for Science, Innovation and Technology. (2025, January 31). Code of practice for the cyber security of AI. UK Government. https://www.gov.uk/government/publications/ai cyber security code of practice/code of practice for the cyber security of ai.
  8. Gale, N. K., Heath, G., Cameron, E., Rashid, S., & Redwood, S. (2013). Using the framework method for the analysis of qualitative data in multi disciplinary health research. BMC Medical Research Methodology, 13, 117. https://doi.org/10.1186/1471 2288 13 117.
  9. Greshake, K., Abdelnabi, S., Mishra, S., Endres, C., Holz, T., & Fritz, M. (2023). Not what you’ve signed up for: Compromising real world LLM integrated applications with indirect prompt injection. arXiv preprint arXiv:2302.12173. https://doi.org/10.48550/arXiv.2302.12173.
  10. Hansen, R., & Venables, P. (2023, June 8). Introducing Google’s Secure AI Framework. Google. https://blog.google/innovation and ai/technology/safety security/introducing googles secure ai framework/.
  11. Microsoft. (2026, August 18). Data, privacy, and security for Microsoft Copilot. Microsoft Learn. Retrieved September 28, 2026, from https://learn.microsoft.com/en us/microsoft 365/copilot/microsoft 365 copilot privacy.
  12. National Cyber Security Centre. (2023, November 27). Guidelines for secure AI system development (Version 1.0). https://www.ncsc.gov.uk/collection/guidelines secure ai system development.
  13. National Institute of Standards and Technology. (2023). Artificial intelligence risk management framework (AI RMF 1.0) (NIST AI 100 1). https://doi.org/10.6028/NIST.AI.100 1.
  14. National Institute of Standards and Technology. (2024). Artificial intelligence risk management framework: Generative artificial intelligence profile (NIST AI 600 1). https://doi.org/10.6028/NIST.AI.600 1.
  15. OWASP. (2025a). LLM01:2025 Prompt injection. OWASP Gen AI Security Project. Retrieved September 28, 2026, from https://genai.owasp.org/llmrisk/llm01 prompt injection/.
  16. OWASP. (2025b). LLM02:2025 Sensitive information disclosure. OWASP Gen AI Security Project. Retrieved September 28, 2026, from https://genai.owasp.org/llmrisk/llm022025 sensitive information disclosure/.
  17. OWASP. (2025c). LLM05:2025 Improper output handling. OWASP Gen AI Security Project. Retrieved September 28, 2026, from https://genai.owasp.org/llmrisk/llm052025 improper output handling/.
  18. OWASP. (2025d). LLM06:2025 Excessive agency. OWASP Gen AI Security Project. Retrieved September 28, 2026, from https://genai.owasp.org/llmrisk/llm062025 excessive agency/.
  19. Raji, I. D., Smart, A., White, R. N., Mitchell, M., Gebru, T., Hutchinson, B., Smith Loud, J., Theron, D., & Barnes, P. (2020). Closing the AI accountability gap: Defining an end to end framework for internal algorithmic auditing. In Proceedings of the 2020 Conference on Fairness, Accountability, and Transparency (pp. 33–44). https://doi.org/10.1145/3351095.3372873.
  20. Selbst, A. D., boyd, d., Friedler, S. A., Venkatasubramanian, S., & Vertesi, J. (2019). Fairness and abstraction in sociotechnical systems. In Proceedings of the Conference on Fairness, Accountability, and Transparency (pp. 59–68). https://doi.org/10.1145/3287560.3287598.
© 2026 PUB · Seek truth with rigour