Can Tutoring Save Us? AI, Accountability, and the Learning We Owe Students

Tutoring has long been one of education’s most powerful and most elusive levers for improving student outcomes. Decades of research show that high-quality tutoring can dramatically accelerate learning. And yet, at scale, that promise has repeatedly fallen short. Systems haven’t sustained it. Students haven’t seen the gains we hoped for. And too often, tutoring has remained fragile, episodic, or disconnected from core instruction.Now AI has entered the equation...not as a silver bullet, but as a forcing function. AI can diagnose faster, personalize more precisely, and deliver feedback at scale. But it also makes one reality impossible to ignore: technology can’t fix weak instructional design or unclear accountability. In an AI-enabled world, the question isn’t whether tutoring can work — it’s whether we are finally willing to build it in ways that actually deliver learning.In this session we'll explore whether AI can meaningfully strengthen tutoring, or does it simply expose what was never built to last? What conditions make tutoring durable? And what does accountability look like when the goal isn’t participation…but real gains for students?