Seven MIT Faculty Members and Alumni Named to Prestigious 2026 AI2050 Fellows Cohort for Pioneering Beneficial Artificial Intelligence Research

The landscape of artificial intelligence research received a significant boost as Schmidt Sciences announced its distinguished 2026 cohort of AI2050 Fellows. Among the honorees are five current Massachusetts Institute of Technology (MIT) faculty members and two prominent alumni, highlighting the institution’s enduring role at the forefront of technological innovation. This year’s selection underscores a collective global push to ensure that rapid advancements in machine learning and algorithmic systems translate into equitable, safe, and profoundly positive outcomes for society over the coming decades.

Conceived and co-chaired by Eric Schmidt and James Manyika, AI2050 operates as a key philanthropic initiative under Schmidt Sciences. The program is specifically structured to address and solve complex, "hard problems" in artificial intelligence. Rather than focusing solely on incremental capability gains, AI2050 challenges its fellows to work backward from a hypothetical future in the year 2050. The guiding question posed to every researcher in the cohort is both visionary and pragmatic: It is 2050. AI has turned out to be hugely beneficial to society. What happened? What are the most important problems we solved and the opportunities and possibilities we realized to ensure this outcome?

By grounding their research in this affirmative future-casting, the fellows are tasked with mitigating the existential, ethical, and practical risks associated with advanced computational systems while amplifying their societal utility. The inclusion of seven individuals tied to MIT speaks to the institute’s multidisciplinary ecosystem, spanning economics, computer science, mathematics, political science, biology, and management.

Leadership and Vision: The AI2050 Senior Fellow

Leading the MIT contingent in the 2026 cohort is Sendhil Mullainathan, who has been named an AI2050 Senior Fellow. Mullainathan holds the Peter de Florez Professor chair at MIT, maintaining dual appointments in the Department of Economics and the Department of Electrical Engineering and Computer Science (EECS), alongside his role as a principal investigator at the MIT Laboratory for Information and Decision Systems (LIDS).

Mullainathan’s current scholarly trajectory sits at the critical intersection of algorithms and human behavior. His focal point is the launch of The Bike Shop@MIT, an innovative research center dedicated to developing human-centered AI through scalable "bicycles for the mind" and novel architectural gears for machine learning algorithms. His extensive background includes co-founding ideas42, a nonprofit applying behavioral science to social problems; MIT’s Abdul Latif Jameel Poverty Action Lab (J-PAL), which champions randomized controlled trials in economic development; and Dandelion Health, a venture blending proprietary AI tools with real-world patient data.

Recognized globally for his contributions—including a MacArthur "Genius Grant," designations as a World Economic Forum Young Global Leader, and inclusion in Foreign Policy Magazine’s Top 100 Thinkers—Mullainathan’s AI2050 project centers on translating complex patterns detectable by AI into rigorous, testable scientific theories. His work addresses a fundamental limitation in contemporary machine learning: while models excel at pattern recognition, they frequently fail to provide explanatory frameworks that human scientists can validate or falsify.

Early Career Innovators Shaping the Next Quarter-Century

In addition to Mullainathan’s senior fellowship, four early-career MIT faculty members have been selected for the 2026 cohort, representing diverse disciplines within the MIT Schwarzman College of Computing and partner departments.

Alex Damian, an assistant professor in the Department of Mathematics with a shared position in EECS and LIDS, joins the cohort to focus on effective AI model training. Damian’s core research investigates the mathematical foundations of deep learning, placing special emphasis on optimization dynamics and representation learning. Before joining MIT, Damian served as a research fellow at the Kempner Institute at Harvard University, earning his PhD in applied and computational mathematics from Princeton University and his BS in mathematics from Duke University. His AI2050 project directly addresses the computational and theoretical bottlenecks in training modern neural networks more efficiently and transparently.

Bailey Flanigan brings expertise at the intersection of computer science, game theory, and democratic governance. As the Theodore T. Miller (1922) Career Development Assistant Professor, Flanigan holds a shared appointment across EECS, Political Science, and LIDS. Having completed her PhD in computer science at Carnegie Mellon University and a fellowship at the Harvard Ash Center for Democratic Innovation, Flanigan’s research designs tools—such as advanced sampling algorithms and preference elicitation methods—to strengthen public participation in democracy. Her AI2050 project centers on gathering structured, actionable public input that lawmakers can reliably utilize when navigating complex legislative decisions.

Yunha Hwang, the Samuel A. Goldblitch Career Development Assistant Professor, bridges the gap between computation and the life sciences with shared appointments in EECS, Biology, and LIDS. Hwang is also the co-founder and chief scientist of Tatta Bio, a scientific nonprofit advancing genomic AI for biological discovery. With a PhD in biology from Harvard University and a BS in computer science from Stanford University, Hwang focuses her research on machine learning for microbial evolution, open science, and biomanufacturing. Her AI2050 project tackles the formidable challenge of determining biological enzyme functions, an advancement that could radically accelerate drug discovery and synthetic biology.

Manish Raghavan rounds out the faculty cohort as the Drew Houston (2005) Career Development Professor at the MIT Sloan School of Management, with a shared appointment in EECS and LIDS. Following postdoctoral research at the Harvard Center for Research on Computation and Society, Raghavan’s work applies computational techniques to matters of profound social concern, including algorithmic fairness and behavioral economics. His research frequently scrutinizes the integration

of automated tools within high-stakes environments like hiring pipelines. His AI2050 project focuses on mitigating the negative externalities of widespread societal reliance on AI while fortifying its constructive aspects.

Beyond the core faculty, MIT’s broader academic ecosystem is further represented by alumni Irene Chen (PhD ’22) and Allison Z. Koenecke (’14), alongside former postdoc Juan Carlos Perdomo, who were likewise named AI2050 Early Career Fellows.

Chronology and Background of the Initiative

The AI2050 initiative was formally established by Schmidt Sciences to bridge a critical gap in contemporary technology funding. While vast sums of capital are funneled into commercial product development and raw compute scaling, long-term foundational research addressing the societal risks and guardrails of AI has historically remained under-resourced.

The program operates on a multi-year funding and fellowship model designed to give top-tier interdisciplinary researchers the financial autonomy and collaborative network necessary to pursue high-risk, high-reward inquiries. By convening cohorts annually, Schmidt Sciences has steadily built an international network of scholars tasked with moving the needle on governance, safety, economic integration, and technical alignment. The selection of the 2026 cohort follows a rigorous global review process that evaluates candidates not only on their technical excellence but also on the societal relevance and ambition of their proposed projects.

Broader Implications and Analytical Outlook

The heavy representation of MIT researchers in the 2026 AI2050 cohort highlights several broader trends in the evolution of artificial intelligence research. First, it signals an institutional shift away from pure computer science isolationism. Every selected MIT fellow maintains cross-departmental appointments or interdisciplinary affiliations—linking computer science with economics, mathematics, political science, biology, and management. This structural convergence reflects a growing consensus that the challenges posed by AI—ranging from algorithmic bias and democratic erosion to genomic discovery and economic dislocation—cannot be solved by technical optimization alone.

Second, the projects funded within this cohort directly target the primary points of friction between artificial intelligence systems and human institutions. Whether through Mullainathan’s push for interpretable scientific theories, Flanigan’s work on algorithmic democratic engagement, or Raghavan’s analysis of labor pipelines, the research seeks to build guardrails before deployment scale creates systemic vulnerabilities.

As artificial intelligence systems continue to embed themselves into the foundational infrastructure of global society, initiatives like AI2050 serve as a vital counterweight to short-term commercial pressures. By supporting researchers who dare to define what a successfully managed technological future looks like in 2050, Schmidt Sciences and institutions like MIT are laying the groundwork for a technological paradigm that prioritizes human well-being, equity, and scientific truth.