HBS AI Institute
Confidence 0.75 · 3 sources · last confirmed 2026-10-15
The Harvard Business School AI Institute, a research institute at HBS co-founded and chaired by Karim Lakhani, with Jen Stave as executive director. Its site publishes short research summaries of work by HBS faculty and associates, bylined HBS AI Content & Learning. With Microsoft it runs the Frontier Firm AI Initiative, in which member companies run AI experiments with Harvard and Microsoft researchers, starting with sales, IT and finance, and attend C-suite programmes. Raffaella Sadun is a research leader in the initiative.
The wiki’s existing Harvard pages call Lakhani’s institute the Digital Data Design Institute at Harvard (D^3). Staff photos on the AI Institute’s site are still filed under that name, and Lakhani’s bio now calls him co-founder and faculty chair of the HBS AI Institute. That suggests the institute was renamed, but no source in the corpus says so. Treat the identification as likely, not confirmed.
Wiki contributions
- 2026-10-15-hbs-ai-institute-frontier-firm-ai-initiative-about — the Frontier Firm definition and four early insights from a forthcoming case: structured experimentation, the managerial problems of running it, the Skill–Will 2×2, and agents as a sharpener of winners and losers.
- 2026-09-03-hbs-ai-institute-think-youre-a-good-leader-try-leading-an-ai-team — summary of Weidmann, Xu & Deming, adding the proposal that AI teams become a leadership practice environment.
- 2025-08-05-hbs-ai-institute-mastering-change-resilience-key-to-ai-driven-success — summary of Lakhani, Stave, Ng & Martines’s HBR article: sensing, rewiring, lock-in; Learn–Do–Imagine–Act–Care.
Why it is an entity (promotion note)
Its byline appears in author: on two source pages in the same ingest, and it hosts the initiative behind a third, so the second-source promotion rule is met on creation.
Reading its sources
The institute summarises work by its own faculty, associates and leadership, and runs the Frontier Firm initiative with a vendor. Its pages are accurate digests but not independent. Cite the underlying paper or article where the wiki holds it.