Agentic AI Deployment in Infrastructure-limited Environments: Observability Gaps, Failure Modes, and AI Governance Primitives

Abstract

This paper discusses the application of agentic Artificial Intelligence (AI) systems to infrastructure-constrained environments and observability gaps, failure modes, and AI governance primitives, in particular. The study measures system performance in a range of different resource setups, finding that observability coverage degrades dramatically by 31 points (61% in low-resource setups versus 92% in high-resource setups), implying a 31-percent decrease in monitoring capacity. The limitation also increases the number of failures of the system between 12 and 34 percent in meaning that there is nearly triple the instability in the operation. The most common failure modes in the analysis are data loss and model drift, with the two factors representing 58% of the total number of failure modes.

The performance of enhanced observability frameworks increases their accuracy of detecting anomalies by 11% with 93% accuracy as compared to 82% accuracy. Moreover, the use of AI governance primitives like enforcement of policies and fallback mechanisms increase the compliance rates to 89 percent and error recovery rates to 83 percent. Latency analysis shows that the delays in decision-making are reduced by 55% in constrained conditions; edge-based processing reduces the latency as well as enhances system performance by approximately 27% under constrained conditions.

In general, the joint implementation of observability, failure detection, and governance mechanisms lead to an increase in the performance of the system between 18% and 30%. The findings provide an insight into the necessity of having the integrated, light and scaled AI systems to ensure reliable, transparent and effective execution under resource limited environment.

REFERENCES

[1] N. Kumar, X. Wei, and H. Zhang, "Agentic Artificial Intelligence as a New Frontier in Information Systems: Promise, Peril, and Research Opportunities," Information & Management, p. 104317, 2026.

[2] D. T. Bansal and R. Gupta, AI in Teacher Education. Geh Press, 2026.

[3] C. Manouan, M. Anagbah, and J. Barros, "Take the Train: Africa at the Crossroad of Modern AI," arXiv preprint arXiv:2603.21795, 2026.

[4] S. P. Dhole, "Latest Trends and Defensive Strategies in Cybersecurity: Emerging Threats, Research Gaps, and Future Directions," 2025.

[5] R. P. Madnani, "An Empirical Study on Assessing the Trustworthiness of AI Systems in Handling Sensitive Financial Information," Rethink Revolution: Transformative Waves Across Disciplines, vol. 80, 2024.

[6] R. Islam Afraim, "Agentic AI as Coordination Infrastructure Technology: Structural Implications for Firms, Growth, and Economic Divergence," Growth, and Economic Divergence, Feb. 2026.

[7] A. Warrier and K. S. Abhilash, "Autonomous Agentic AI for Clinical Workflow Orchestration: Self-Managing Healthcare Operations," in Proc. 2025 6th Int. Conf. IoT Based Control Networks and Intelligent Systems (ICICNIS), Dec. 2025, pp. 1450–1457.

[8] M. A. Islam, M. Rahman, F. Dal Mas, S. E. Haque, and U. Hani, "Navigating institutional and capability barriers in agentic artificial intelligence adoption: evidence from small and medium enterprises in Bangladesh," VINE Journal of Information and Knowledge Management Systems, pp. 1–25, 2026.

[9] S. Ahmed and M. Asif, "Comparative analysis of attitudes toward climate change policies across urban and rural populations," Pakistan Journal of Social Science Review, vol. 5, no. 1, pp. 747–769, 2026, doi: 10.5281/zenodo.18457821.

[10] S. Ahmed and M. Asif, "Public opinion on the effectiveness of local government anti-corruption measures: A multi-city survey analysis," International Journal of Social Sciences Bulletin, vol. 4, no. 1, pp. 1189–1201, 2026, doi: 10.5281/zenodo.18412790.

[11] M. Asif and S. Ullah, "Determinants of support for federalism vs. centralization: A survey of public opinion in Punjab and Khyber Pakhtunkhwa (KP)," Social Science Review Archives, vol. 4, no. 1, pp. 2791–2807, 2026, doi: 10.70670/sra.v4i1.1843.

[12] M. Asif and S. Ullah, "Performance voting vs. identity voting: An analysis of electoral behaviour in Pakistani districts," Journal of Applied Linguistics and TESOL (JALT), vol. 9, no. 1, pp. 213–226, 2026, doi: 10.63878/cjssr.v4i1.2079.

[13] M. Asif, A. Ali, and F. A. Shaheen, "Assessing the effects of artificial intelligence in revolutionizing human resource management: A systematic review," Social Science Review Archives, vol. 3, no. 4, pp. 2887–2908, 2025, doi: 10.70670/sra.v3i3.1055.

[14] M. Asif and R. J. Asghar, "Managerial accounting as a driver of financial performance and sustainability in small and medium enterprises in Pakistan," Center for Management Science Research, vol. 3, no. 7, pp. 150–163, 2025, doi: 10.5281/zenodo.17596478.

[15] M. Asif, M. A. Pasha, and A. Shahid, "Energy scarcity and economic stagnation in Pakistan," Bahria University Journal of Management & Technology, vol. 8, no. 1, pp. 141–157, 2025.

[16] M. Asif, H. Shah, and H. A. H. Asim, "Cybersecurity and audit resilience in digital finance: Global insights and the Pakistani context," Journal of Asian Development Studies, vol. 14, no. 3, pp. 560–573, 2025, doi: 10.62345/jads.2025.14.3.47.

[17] V. V. R. M. Bokka, "Architecting Agentic AI for Autonomous Data Protection in Enterprise Systems," Journal of Engineering and Computer Sciences, vol. 4, no. 12, pp. 47–55, 2025.

[18] A. Coelho, P. Ribeiro, H. Fontes, and R. Campos, "A4FN: an Agentic AI Architecture for Autonomous Flying Networks," in Proc. 2025 IEEE 36th Int. Symp. Personal, Indoor and Mobile Radio Communications (PIMRC), Sep. 2025, pp. 1–6.

[19] S. Ahmed and M. Bashir, "Middle East Conflicts & Their Influence on International Politics," Inverge Journal of Social Sciences, vol. 5, no. 1, pp. 160–172, 2026, doi: 10.63544/ijss.v5i1.228.

[20] M. Bashir and S. Ullah, "Public Trust in State Institutions: A Comparative Analysis of the Military, Judiciary, and Parliament in Post-2023 Election Pakistan," Inverge Journal of Social Sciences, vol. 5, no. 2, pp. 88–100, 2026, doi: 10.63544/ijss.v5i2.248.

[21] S. Joshi, "A Comprehensive National Framework for Federal AI Adoption: Governance, Standards, and Implementation Pathways for Agentic and Generative Systems," Standards, and Implementation Pathways for Agentic and Generative Systems, Feb. 2026.

[22] T. Ramkissoon, A. Ali, D. Ramsubhag, L. Addison, and P. Hosein, "A Policy Framework for AI-Centric Climate Resilience Action Plans," in Climate Action and Sustainable Development Goals, CRC Press, 2026, pp. 145–162.

[23] L. Addison, K. Rajaram, A. Hosang, T. Tewari, T. A. Tuitt, and P. Hosein, "An Ai-Driven Early Warning and Forecast-Based Framework for Disaster Resilience in Small Island States," Available at SSRN 5381918.

[24] A. Almusaed, A. Almsaad, and I. Yitmen, "Smart Cities That Think: Cognitive Infrastructures, AI Governance, and Sustainable Urban Futures," in Smart Cities-Designing the Future of Urban Living, IntechOpen, 2025.

[25] Z. Lyu et al., "Empowering intelligent low-altitude economy with large AI model deployment," IEEE Wireless Communications, 2025.

[26] A. Baruwa, "AI powered infrastructure efficiency: enhancing US transportation networks for a sustainable future," International Journal of Engineering Technology Research & Management (IJETRM), vol. 7, no. 12, pp. 329–350, Dec. 2023.

[27] E. Goodluck, "Smart Rural Infrastructure Planning Using Machine Learning and Data-Driven Optimization Models."

[28] D. Ferreira, P. Ribeiro, A. Coelho, and R. Campos, "Design and Evaluation of a Multi-Agent Perception System for Autonomous Flying Networks," arXiv preprint arXiv:2512.00259, 2025.

[29] S. Calado, "Educators as Cognitive Architects in AI-Augmented Learning Environments: Redefining Professional Agency and Pedagogical Co-Performance," in Human-AI Collaboration and the Evolving Role of Educators, IGI Global Scientific Publishing, 2026, pp. 89–122.

[30] L. Jia and Y. Pei, "Recent advances in multi-agent reinforcement learning for intelligent automation and control of water environment systems," Machines, vol. 13, no. 6, p. 503, 2025.

[31] D. Mohiuddin, “Adaptive marketing systems and consumer feedback loops: Implications for market development in emerging economies,” Journal of Business Insight and Innovation, vol. 5, no. 1, pp. 37–48, 2026.

[32] D. Mohiuddin, “HR tech adoption in digital banking: Implications for workforce development and financial sector growth in emerging economies,” Journal of Business Insight and Innovation, vol. 4, no. 2, pp. 77–90, 2025.

[33] D. Mohiuddin and D. N. Farhan, “Artificial intelligence in marketing: Ethical challenges and solutions for consumers and society,” Journal of Business Insight and Innovation, vol. 4, no. 1, pp. 73–87, 2025.

[34] D. Mohiuddin, “Algorithmic hyper-personalization: The double-edged sword of predictive personalization—An empirical investigation,” Journal of Engineering and Computational Intelligence Review, vol. 2, no. 2, pp. 82–94, 2024.

[35] D. Mohiuddin, “Consumer perceptions and trust in AI-generated advertising: An experimental study in the Pakistani context,” Apex Journal of Social Sciences, vol. 3, no. 1, pp. 53–68, 2024.