At the Berkman Klein Center, 2025 was a year of gathering momentum and crystallizing focus, aligning people, programs, and questions around a shared direction.
Last summer, BKC welcomed Executive Director Alex Pascal, who brought his experience in technology governance, domestic policy, and national security to the Center. Since his arrival, Alex has animated the Center with new direction, energy, and focus on generating impact in the public interest. Josh Joseph also joined BKC over the summer as Chief AI Scientist after co-teaching “Agentic AI and the Law” with BKC Faculty Director Jonathan Zittrain. In his first months, he has built up BKC’s in-house AI research infrastructure to lay the foundation for a strong 2026 and launched numerous projects, including efforts to measure and constrain the agency of AI systems and to develop benchmarks that move beyond narrow notions of “intelligence.” The Applied Social Media Lab launched new public interest products, and its precursor, the Institute for Rebooting Social Media, concluded its three-year cycle. We brought BKC’s transparency, privacy, and safety projects onto international stages, from Barcelona to Geneva, and advanced the Center’s work stateside, from Silicon Valley to the Massachusetts State House.
Opening the Black Box: BKC’s AI Interpretability Workshop
BKC put its role as a trusted convener in responsible AI on display at an AI Interpretability Workshop led by Faculty Associates Fernanda Viégas and Martin Wattenberg, in collaboration with Faculty Director Jonathan Zittrain. The workshop brought together technologists, researchers, and practitioners to explore the policy implications of “TalkTuner,” an interpretability dashboard that visualizes—in real time—how a chatbot infers aspects of a user’s identity, such as age or gender, and how those inferences can change a chatbot’s responses to a user.
The conversations surfaced many of the core questions behind BKC’s interpretability work, such as where might the benefits of transparency be outweighed by the risks of reinforcing bias or enabling misuse? The workshop exemplified the Center’s broader aim to connect cutting-edge technical research with the ethical, institutional, policy, and legal challenges such research raises.
Partnering for a Unique Digital Policy Gathering
In June, BKC partnered with the leading membership organization for privacy professionals, the IAPP, to host the inaugural Navigate: Digital Policy Leadership Retreat. The two-day gathering convened global leaders from academia, government, industry, and civil society to engage deeply with digital policy challenges, including AI governance, cybersecurity, online safety, and technological sovereignty. The retreat demonstrated the value of sustained, cross-sector engagement and marked an important step in BKC’s expanding professional learning and convening work.
Advancing the Debate: BKC’s 2025 Speaker Series on AI
In 2025, the Berkman Klein Center convened vital, multi-perspective conversations on artificial intelligence through a dynamic speaker series that featured leading scholars, technologists, and policymakers with new ideas. In the spring of 2025, sessions ranged from “How Do AI Chatbots See Us?” (Fernanda Viégas and Martin Wattenberg) to “Open Source Lawfare: AI Regulation After DeepSeek” (Ben Brooks) to “Legal Frameworks for Governing AI Agents” (Noam Kolt). In the fall, we built on that foundation by focusing more sharply on AI policy frameworks and competing visions of AI’s future, with events including “Belief, Uncertainty, and Truth in Language Models” (Jacob Andreas) and the critical panel “AI Triad: A Dialogue Across Differences” (Jason Crawford, Amba Kak, Brian McGrail), the latter facilitated by Faculty Director Jonathan Zittrain. Together, these series cemented BKC’s place as a forum where complex questions about AI design, governance, and impact are raised and debated within a big tent, with curiosity, open-mindedness, and deep concern for people.
Sunsetting the Institute for Rebooting Social Media
In 2025, the Berkman Klein Center concluded the Institute for Rebooting Social Media, a three-year “pop-up” launched in 2021 to rethink how social technologies might better serve the public good. Emerging during a period of profound social and political disruption, the Institute assembled an interdisciplinary community of nearly 40 scholars and practitioners whose work examined the promises and perils of the digital public sphere. Its legacy continues through the Applied Social Media Lab and new BKC initiatives focused on accountability, public discourse, and democratic resilience.
Making Sense of AI’s Expanding Frontiers:
An Interview with Josh Joseph, BKC’s Chief AI Scientist
Built with Shorthand