Karen Hao explores power, accountability, and the future of AI
As part of the CBC Ideas series, the Schwartz Reisman Institute welcomed Karen Hao to the University of Toronto to discuss the political economy of AI development, the need for stronger accountability, and the importance of building alternative, less resource-intensive approaches to AI systems.
On March 11, 2026, the Schwartz Reisman Institute for Technology and Society (SRI) welcomed award-winning journalist and author Karen Hao to the University of Toronto for a day of events examining the global political economy of artificial intelligence.
Hao, author of the best-selling book Empire of AI: Dreams and Nightmares in Sam Altman’s OpenAI (2025), engaged with undergraduates, graduate students, and faculty across two sold-out events—marking her first visit to U of T.
Hao’s visit comes at a moment of critical debate over how AI systems should be governed and who they should serve. As governments, industry leaders, and civil society grapple with questions of regulation, safety, and economic disruption, her work offers a critical lens on the structural forces shaping the field, foregrounding issues of power, resource allocation, and public accountability.
This event is part of the CBC series, “Ideas with Nahlah Ayed”. Listen now on CBC, or through your preferred podcast app.
Rethinking AI development and accountability
The evening keynote at the Rotman School of Management, moderated by Nathalie A. Smuha, brought a full audience into conversation on power, governance, and the global impacts of AI development.
Drawing on her investigative reporting, Hao examined how today’s dominant AI systems are shaped by a particular model of innovation—one driven by scale, capital concentration, and resource intensity. She argued that this paradigm not only consolidates power among a small number of firms but also narrows the range of technological futures that are being actively pursued.
A central theme of the talk was the need to distinguish between different kinds of AI systems and their appropriate uses. Hao cautioned against treating all AI as a single category, suggesting that doing so obscures meaningful differences in cost, impact, and purpose. She likened this to grouping cars, trains, rockets, and bicycles under one category simply called “transportation”—a framing that makes it harder to ask whether societies should be investing in the infrastructure to build better rockets or trains.
The evening keynote at the Rotman School of Management, moderated by Nathalie A. Smuha (left) and special guest, award-winning journalist and author Karen Hao.
Extending this analogy, Hao called for greater attention to what might be considered the “bicycles” of AI: systems that are more efficient, accessible, and aligned with public needs, rather than dependent on massive computational infrastructure and energy consumption. In contrast to the current scaling paradigm underpinning large language models, Hao explored how alternative approaches could reduce resource demands while expanding the range of actors able to participate in AI development, and shared compelling stories of the human cost of the current development paradigm which can exploit workers and enact a costly toll on our environment.
Throughout the discussion, Hao emphasized that questions of accountability are inseparable from technical design choices. Who builds AI systems, how they are funded, and what incentives guide their development all shape their societal impact. Addressing these issues, she argued, requires stronger forms of public-interest governance alongside new models of innovation.
A masterclass on storytelling and resistance
Earlier that afternoon, SRI hosted an invitation-only masterclass led by Hao in conversation with Anna Su at U of T’s Henry R. Jackman Faculty of Law. The session offered students and faculty the opportunity to engage directly with Hao’s approach to reporting on AI and to reflect on how research and journalism can respond to concentrated technological power.
Hao spoke about her transition from engineering to journalism, describing a growing awareness of the gap between the ideals of technological progress and its real-world consequences. “If you want to make this better,” she noted, “you have to influence the people actually making it.”
Participants raised questions on topics ranging from nonprofit governance and labour impacts to AI sovereignty and democratic accountability. Across these exchanges, Hao emphasized the importance of both critique and construction—challenging existing systems while also building viable alternatives.
SRI hosted an invitation-only masterclass led by Karen Hao (left) in conversation with SRI Research Lead, Anna Su, at U of T’s Henry R. Jackman Faculty of Law.
She also reflected on the role of narrative in shaping public understanding. “To write a book with broad appeal you need characters,” she explained, describing how Empire of AI uses individual stories to illuminate broader structural dynamics. “Through telling the story through people you begin to see the banality of the empire.”
Despite the scale of the challenges she documents, Hao pointed to emerging forms of resistance as a source of momentum. “Every single person can engage in an act of resistance, and build an alternative,” she said. “That is something you can do tomorrow that would have a huge impact.” Hao also observed that U of T is “perfectly placed” as a site for the kinds of AI research and development the world needs right now, noting the university’s considerable talent and its potential to act as a hub convening diverse stakeholders from different disciplines and backgrounds.
Empire of AI: Dreams and Nightmares in Sam Altman’s OpenAI (Penguin, 2025).
Building alternative futures
Across both events, a shared message emerged: the trajectory of AI is not fixed. While current systems reflect a particular set of economic and political priorities, alternative paths remain possible.
The day’s conversations highlighted the role that universities, researchers, and students can play in shaping those paths—whether through critical scholarship, public engagement, or the development of new technological approaches that better align with societal values.
“Karen’s visit to the University of Toronto created a rare and valuable space for students and researchers to engage directly with someone who is shaping global conversations on AI accountability,” says Su. “It was especially inspiring to see the depth of curiosity and critical engagement from the next generation—these are exactly the kinds of interdisciplinary perspectives we need to meet the challenges of today’s AI landscape.”
As AI systems become more deeply embedded in everyday life, questions of accountability, design, and collective choice will only become more urgent, highlighting the need for sustained interdisciplinary research and dialogue around how these technologies are developed and governed.
Want to learn more?
Browse stories by tag:
- AI
- AI safety
- AI trust
- Computer science
- Computer security
- Copyright
- Cybersecurity
- Data
- Democracy
- Economics
- Education
- Engineering
- Ethics
- GPO-AI
- Governance
- Health
- Human Rights
- In the Media
- Jobs
- LLMs
- Law
- Normativity
- Philosophy
- Political Science
- Privacy
- Privacy Series
- Psychology
- Public Policy
- Recommenders
- Regulation
- Religion
- Reports
- Trust
- Workshops
