As 2026 unfolds, the future of research and organizational decision-making cannot be understood as an AI story alone. Research ecosystems are under strain from declining trust, growing precarity, and expanding governance demands, and AI often amplifies these pressures rather than resolving them. The trends that will matter most are structural: how knowledge is governed, how methodological authority is sustained, and how accountability and coordination are maintained across research and risk management in an increasingly complex environment.
By Silvana di Gregorio, PhD
Product Research Director and Head of Qualitative Research
As 2026 gets underway, it is tempting to frame the future of research and organizational decision-making as an “AI moment”. AI certainly matters—but focusing on AI alone obscures deeper forces reshaping how knowledge is produced, taught, governed, and acted upon.
At Lumivero, we believe real progress happens when innovation is grounded in reflection—when new capabilities build on proven methods rather than replace them. Research and risk management have long served as engines of progress, helping institutions and organizations turn complexity into insight. AI expands what’s possible by enabling work with broader, more diverse datasets and allowing uncertainty to be explored at scale. Yet technology on its own does not create clarity. Without trusted methodologies, shared understanding, and accountable governance, new tools risk amplifying existing challenges rather than resolving them.
Those challenges are already significant. Research ecosystems are under strain as public trust in higher education and science has declined markedly over the past decade, with confidence in universities and research institutions eroding across multiple countries and political contexts (Gallup & Lumina Foundation, 2023) although a 2025 poll showed a modest reversal in that trend (Gallup & Lumina Foundation, 2025). At the same time, academic labor markets are increasingly characterized by short-term contracts, limited career progression, and structural precarity, particularly for early-career researchers (OECD, 2024).
At the same time, governance demands have expanded dramatically. Research systems now operate under layered regimes of ethics oversight, data protection, compliance reporting, performance monitoring, and audit—placing increasing administrative and cognitive burden on institutions and researchers alike (European Commission, 2024).
AI is entering into this instability—not resolving it. In many cases, it intensifies existing pressures by accelerating expectations of productivity, scale, and standardization without addressing underlying issues of trust, capacity, or legitimacy.
The trends that will matter most in 2026 are therefore not technological breakthroughs in isolation, but structural shifts in methodological authority, coordination, governance, and accountability across research and risk management.
Across academia, policy research, and industry, researchers are being asked to produce insights faster, at scale, and under increasing pressure to standardize outputs. At the same time, long-standing methodological traditions—particularly qualitative, interpretive, and reflexive approaches—are increasingly required to justify their legitimacy in metric-driven environments.
This tension surfaced sharply in 2025 with the publication of an open letter signed by more than 400 qualitative researchers rejecting the use of generative AI in reflexive qualitative research (Jowsey et al., 2025). The letter does not reject technology wholesale; it defends meaning-making, reflexivity, and epistemic accountability as irreducibly human practices that cannot be automated without loss.
A subsequent rebuttal (Friese, 2025) argues for a more differentiated stance, positioning AI as a potential analytic scaffold rather than a substitute for interpretation. What matters for 2026 is not which position prevails, but that methodological boundaries are now being articulated explicitly, rather than quietly eroded through convenience or institutional pressure.
Prediction: In 2026, methodological resistance will become more visible and influential, shaping norms of credible research across academic and applied contexts.
A quieter but significant shift is underway from AI tools that merely respond to prompts toward agentic systems that can plan, act, monitor outcomes, and iterate with limited human intervention.
In research environments, agentic systems are being explored for literature discovery, data preparation, workflow orchestration, and iterative analysis (Wu et al. 2025; Jaradeh and Auer.,2025; Xu et al., 2025).
In risk management, there have been efforts to harness the power of agentic systems for tasks relevant to the financial services industry (Okpala et al., 2025) as well as a study that explored using multi-agent systems for real-time risk management in a large-scale logistics company, integrating AI and sociotechnical systems theory for efficiency. (Bonrath & Eulerich, 2025).
These systems are at very early stages and don’t replace people but are increasingly embedded to handle routine risk tasks and alert humans when something needs attention, helping organizations manage risks more quickly and clearly (Jackson, 2025).
While these systems promise efficiency and coordination, they also redistribute responsibility in subtle ways. When systems decide what to do next, rather than merely what to suggest, questions of accountability, traceability, and error attribution become harder—not easier—to answer.
Prediction: By 2026, agentic systems will increasingly underpin research platforms and risk infrastructures, intensifying the need for governance and human oversight rather than reducing it.
As uncertainty grows more interconnected, the challenge for risk professionals is shifting from analysis to alignment. Many organizations already generate sophisticated risk models, scenarios, and registers—but struggle to translate those insights into shared understanding across leadership teams. Research on risk governance increasingly emphasizes that communication is not a secondary activity, but a core mechanism through which organizations make sense of complex, uncertain risk environments and coordinate action (Renn & Benighaus, 2024).
Risk information is often technically sound but poorly communicated. Static reports, disconnected dashboards, and siloed risk views make it difficult for decision-makers to see how risks interact, evolve, or compound across programs and portfolios. The result is not a lack of data, but a lack of narrative—leaders talk past one another, debate assumptions instead of implications, and delay action. As the literature notes, when risk communication fails to support collective sensemaking, even high-quality analysis can lose its practical value (Renn & Benighaus, 2024).
By 2026, effective risk management will increasingly depend on how well risk can be seen, not just measured. Visual models, shared decision environments, and scenario-based storytelling will become essential tools for aligning diverse stakeholders around trade-offs, timing, and tolerance. Communicative approaches that reveal interdependencies and cascading effects—across supply chains, regulatory exposure, funding, or delivery timelines—will be critical for navigating systemic risk.
Prediction: In 2026, organizations that invest in risk communication as a core capability—rather than a downstream reporting function—will make faster, more confident decisions under uncertainty, while those relying on static documentation will continue to struggle with misalignment and delayed response.Shape
In risk management, the emphasis on prediction and optimization is increasingly challenged by the realities of systemic risk and deep uncertainty. Climate instability, cyber risk, and cascading disruptions have exposed the limits of purely model-driven approaches, particularly in complex, interdependent systems (Liu & Renn, 2025; Ciullo et al., 2025).
As a result, organizations are placing greater value on mixed-methods evidence, scenario reasoning, and managerial judgement to complement quantitative models (Crawford & Jabbour, 2024). For risk management software, this signals a shift from predictive scoring toward tools that support scenario exploration, qualitative context, and transparent decision rationale. AI-supported simulations can extend scenario analysis, but they do not resolve ambiguity or replace human responsibility for high-stakes decisions (Shinkle et al., 2025).
Prediction: By 2026, leading risk-management practices will rebalance predictive modeling with organizational learning, scenario reasoning, and transparent human judgement in the face of systemic uncertainty.Shape
Regulatory divergence is an increasingly important constraint on cross-border research and professional practice. The European Commission’s Science, Research and Innovation Performance of the EU 2024 (SRIP 2024) describes a more complex and uncertain EU research environment, with persistent fragmentation across national systems and renewed calls for coordination and harmonization.
AI governance is also diverging across jurisdictions. The EU’s AI Act (2024) introduces harmonized, risk-based rules for AI systems, while the United States relies on voluntary, sector-agnostic frameworks such as NIST’s AI Risk Management Framework (2023). China’s generative AI measures (Cyberspace Administration of China, 2023) emphasize state oversight and public-facing controls. Although international bodies promote shared principles and standards, adoption remains uneven. For researchers and risk professionals working across borders, these differences translate into real compliance and operational friction, as expectations vary by legal and policy regime.
Prediction: In 2026, organizations operating across jurisdictions will treat AI and data governance as a product and operating constraint, not a legal afterthought—rewarding teams that can design compliant workflows across multiple regimes.
The future of research and risk management will not be decided by AI adoption alone.
It will be shaped by how institutions respond to deeper pressures: declining trust, academic precarity, contested methodological legitimacy, agentic autonomy, systemic risk, and fragmented governance landscapes. AI is part of this story, but it is not the whole story.
Used with clear guardrails, AI can reduce administrative workload by triaging large document sets, drafting structured summaries, and flagging patterns for review—provided outputs are traceable to source evidence and routinely checked. The goal isn’t to outsource judgement, but to free human experts to do the work AI cannot: contextual interpretation, ethical reasoning, and accountable decisions.
The most resilient systems in 2026 will be those that remain clear about what should not be automated, where human judgement must remain central, how relational and supervisory capacity is sustained, and how accountability can be sustained across increasingly complex regulatory environments.
Silvana di Gregorio, PhD, Product Research Director and Head of Qualitative Research at Lumivero
Silvana di Gregorio, PhD, is a sociologist and former academic with a PhD in Social Policy from the London School of Economics. She has been training, consulting, and publishing about qualitative data analysis software since 1995. For 16 years, she had her own training and consulting business, SdG Associates. She is author of, “Voice to Text: Automating Transcription” in Vanover, C., Mihas, P., Saldana. J. (Eds.) Analyzing and Interpreting Qualitative Data: After the Interview, Sage Publications, and “Using Web 2.0 tools for Qualitative Analysis” in Hine, C. (Ed.) Virtual Research Methods. Volume 4, Sage Publications, and co-author with Judith Davidson, “Qualitative Research Design for Software Users,” McGraw-Hill, and “Qualitative Research and Technology: In the Midst of a Revolution” in Denzin, N. and Lincoln, Y. (Eds.). Handbook of Qualitative Research (4th Edition), Thousand Oaks: Sage, and co-author with Linda Gilbert and Kristi Jackson, “Tools for Qualitative Analysis” in Spector, J.M., Merrill, M.D., Elen, J. (Eds.) Handbook of Research on Educational Communications and Technology. She is part of the Product Team at Lumivero.
References
AEI. (2022). The roots of public mistrust in science, policy, and academic integrity. American Enterprise Institute.
Bell, A., Borges Dario, A., Klinner, C., Nisbet, G., Penman, M., Storer, D., & Monrouxe, L. (2025). Improving the quality of allied health placements: student, educator and organisational preparedness. Studies in Continuing Education, 47(1), 337-359.
Bonrath, A., & Eulerich, M. (2025). From Data to Decisions: Real-Time Risk Management Using Multi-Agent Systems. SSRN Working Paper, https://doi.org/10.2139/ssrn.5665890
Boud, D., Molloy, E., & Chang, V. (2025). Framing student navigation of feedback on placements. Teaching in Higher Education, 1-17.
Ciullo, A., Franzke, C. L., Scheffran, J., & Sillmann, J. (2025). Climate-driven systemic risk to the sustainable development goals. PLoS Climate, 4(4), e0000564.
Crawford, J., & Jabbour, M. (2024). The relationship between enterprise risk management and managerial judgement in decision‐making: A systematic literature review. International Journal of Management Reviews, 26(1), 110-136.
Cyberspace Administration of China. (2023). Interim Measures for the Management of Generative Artificial Intelligence Services.
European Commission, Directorate-General for Research and Innovation. (2024). Science, Research and Innovation Performance of the EU 2024. European Union.
European Union. (2024). Artificial Intelligence Act. Official Journal of the European Union.
Friese, Susanne, Response to: “We Reject the Use of Generative Artificial Intelligence for Reflexive Qualitative Research.” (November 01, 2025). Available at SSRN: https://ssrn.com/abstract=5690262 or http://dx.doi.org/10.2139/ssrn.5690262
Gallup & Lumina Foundation. (2023). Public confidence in higher education.
Gallup & Lumina Foundation. (2025). The state of higher education, 2025.
Gustavsson, M., & Bivall, A. C. (2025). The challenges clinical supervisors experience when supervising students in the workplace. Higher Education, Skills and Work-Based Learning, 15(7), 127-138.
Jackson, F. (2025). Governing Autonomous AI Agents with Policy-as-Code: A Multi-Layer Architecture for Risk, Compliance, and Zero-Trust Control. SSRN Working Paper (abstract_id = 5820262).
Jaradeh, M. Y., & Auer, S. (2025). Deep Research in the Era of Agentic AI: Requirements and Limitations for Scholarly Research. In 5th International Workshop on Scientific Knowledge: Representation, Discovery, and Assessment (Sci-K), Nara, Japan (Nov 2–6, 2025). CEUR Workshop Proceedings.
Jowsey, T., Braun, V., Clarke, V., Lupton, D., Fine, M., et al. (2025). We reject the use of generative artificial intelligence for reflexive qualitative research. Manuscript commentary.
Liu, H., & Renn, O. (2025). Polycrisis and Systemic Risk: Assessment, Governance, and Communication: H. Liu et al. International Journal of Disaster Risk Science, 1-24.
NIST. (2023). Artificial intelligence risk management framework (AI RMF 1.0). National Institute of Standards and Technology.URL: https://nvlpubs. nist. gov/nistpubs/ai/nist. ai, 100-1.
OECD. (2024). The state of academic careers in OECD countries: An evidence review. OECD Education Policy Perspectives.
Okpala, I., Golgoon, A., & Kannan, A. R. (2025). Agentic AI Systems Applied to Tasks in Financial Services: Modeling and Model Risk Management Crews. arXiv:2502.05439.
Renn, O., & Benighaus, C. (2024). Risk communication revisited: A governance and sensemaking perspective for complex risk environments. Risk Sciences, 2(1), 100026. https://doi.org/10.1016/j.riss.2024.100026
Shinkle, G. A., Gujarati, C., & Sharry, P. (2025). Scenario Analysis in the Ai Era: Redefining Human Involvement. Available at SSRN 5239542.
Wu, S., Ma, X., Luo, D., Li, L., Shi, X., Chang, X., Lin, X., Luo, R., Pei, C., Du, C., Zhao, Z.-J., et al. (2025). Automated literature research and review-generation method based on large language models. National Science Review, 12(6), nwaf169. https://doi.org/10.1093/nsr/nwaf169
Xu, Q., Amjad, N., Giles, G., Cumming, A., Hermesky, D. A., Wen, A., ... & Kim, Y. (2025). A Multi-Agent Large Language Model Framework for Automated Qualitative Analysis. arXiv preprint arXiv:2512.16063.
