Angelo Romasanta is an Assistant Professor in the Department of Operations, Innovation and Data Science at Esade Business School in Barcelona, where he specializes in Data, Analytics, Technology and Artificial Intelligence. He holds a PhD in Innovation Management from Vrije Universiteit Amsterdam, which he earned with a prestigious Marie Curie scholarship. His academic trajectory reflects a rare integration of scientific depth and managerial insight: he also holds an MSc in Chemistry from the University of Barcelona, Bologna and Algarve, grounding his research in the material realities of scientific and technological innovation. This multidisciplinary foundation—spanning the physical sciences, data science and strategy—positions him at the forefront of one of the most urgent questions facing leaders today: how should organizations integrate artificial intelligence into strategic decision-making without surrendering their distinctive judgment?
Romasanta’s research focuses on the impact of AI upon scientific and technological innovation, with particular emphasis on applications such as drug discovery through technologies like AlphaFold. He has contributed to major European projects coordinated by CERN and published extensively on how digital infrastructures and deep tech commercialization reshape innovation ecosystems. His work consistently interrogates the boundary between what can be automated and what demands human cognition—a theme that runs through his most influential contribution to date.
In March 2026, Romasanta—together with co-authors Llewellyn D.W. Thomas and Natalia Levina—published “Researchers Asked LLMs for Strategic Advice. They Got ‘Trendslop’ in Return” in the Harvard Business Review. The study tested leading AI models across thousands of business dilemmas and seven core strategic tensions—differentiation versus commoditization, augmentation versus automation, long-term versus short-term thinking, among others. The findings were stark: LLMs systematically favored trendy, buzzword-aligned strategies regardless of context. They defaulted to the same fashionable recommendations whether the company was a startup or an incumbent, in growth or decline. Prompt engineering and richer contextual information had limited effect. The models, trained on the vast corpus of internet discourse—blogs, LinkedIn posts, business commentary—had absorbed its biases and amplified them with confident, polished prose.
Romasanta named this phenomenon “trendslop”: the propensity of AI to deliver strategy that sounds fluent and current but substitutes context-specific reasoning with the flavor-of-the-month consensus. The term has since entered the business lexicon, recognized by the Cambridge Dictionary and debated in boardrooms and academic circles alike. But trendslop is not merely a linguistic curiosity. It is, as Romasanta has argued, a process failure, not a prompting failure. It reveals that when leaders consult the same frontier models, they receive the same advice—nudged toward strategic mediocrity and convergence rather than distinctiveness and competitive advantage. In his classroom, Romasanta demonstrated this dynamically: students who used AI converged on nearly identical hybrid recommendations, while those who relied on their own cognition produced more varied and contextually grounded answers.
For the Long-Term Strategy Conference 2026 (LTSP26) , Romasanta’s coinage of “trendslop” is foundational. The conference is organized around a single conviction: that strategy processes break down in predictable ways, and that high-leverage interventions are urgently needed. Trendslop crystallizes one of the most insidious breakdowns of the AI era—the erosion of context-sensitive, long-term strategic thinking by tools that optimize for what is popular rather than what is right. Romasanta’s research does not counsel abandonment of AI; rather, it calls for boundaries: using AI to expand options, not make choices; counteracting known biases; remaining alert to shifting biases; and, above all, preserving human judgment at the center of strategic trade-offs. His work challenges leaders to ask not whether AI can generate strategy, but whether the strategy it generates is worth pursuing—and whether the conditions under which it is deployed are conducive to genuine, long-term advantage.
At LTSP26, Romasanta’s voice is essential not only because he named the problem, but because he continues to chart a path beyond it—toward a discipline of strategic thinking that leverages AI without being captured by its biases, and that keeps the long view firmly in focus.