In June 2024, Luan joined Amazon after licensing Adept's technology. Amazon arranged for Adept's investors to be made whole rather than losing their investment. Luan led Amazon's AGI lab, created in December 2024 for "long-term research bets." However, in February 2026, Luan announced his departure after less than two years. He wrote on LinkedIn: "But with AGI so close, I decided to spend 100% of my time on teaching AI systems brand new capabilities" and plans to "cook up something new." The quick departure raises questions about Amazon's AI lab culture.
All receipts on David Luan
David Luan
Luan structured the 2024 Adept-Amazon deal unusually to protect investors. Unlike typical acqui-hires, Amazon didn't buy Adept and didn't directly pay investors. According to Semafor, Adept's co-founders "didn't want their backers to lose money, so they arranged for them to be made whole using funds paid by Amazon." This came after talks with Meta and Microsoft fell through. Adept had raised $415M+ at $1B+ valuation since 2022 but struggled to bring products to market. The deal allowed Amazon to gain talent and technology while Adept continues operating independently.
Share this receipt
Posts with the receipt card attached.
Post this receiptDownload card
David Luan — Politics: MIXED Luan structured the 2024 Adept-Amazon deal unusually to protect investors. Unlike typical acqui-hires, Amazon didn't buy Adept and didn't directly pay investors. According to Se… https://reciepts.tech/stance/1560
More on David Luan
See all →Luan describes AI ethics as "incredibly important" in his career. As VP of Engineering at OpenAI (2017-2020), he oversaw safety and policy teams. At Axon, he established an AI Safety/Ethics Board and blanket ban on facial recognition on body-worn cameras. He created an interdisciplinary board with representation from the EFF, Future of Humanity Institute, NYU Law, and civil liberties organizations to regulate AI in law enforcement, ensuring ML is used "solely to improve transparency and accountability" and prohibiting surveillance or decision-making use. He recently co-authored research on "Chain of Thought Monitorability" for AI safety.