"Not a Paper: “Frontier Lab CEOs are Capable of In-Context Scheming”" by LawrenceC

About this episode

The episode discusses the risks associated with executive misalignment in frontier AI companies and proposes threat models for potential harm.

(Fragments from a research paper that will never be written) Extended Abstract. The frontier AI developers are becoming increasingly powerful and wealthy, significantly increasing their potential for risks. One concern is that of executive misalignment: when the CEO has different incentives and goals than that of the board of directors, or of humanity as a whole. Our work proposes three different threat models, under which executive misalignment can lead to concrete harm. We perform t...

People in this episode

Guest: LawrenceC

Topics covered

• AI development
• executive misalignment
• threat models
• risks of power
• incentives

Keywords

• AI
• CEOs
• executive misalignment
• threat models
• risks

More episodes of LessWrong (Curated & Popular)

• "PSA: Almost nobody is working on alignment" by Chi Nguyen, peterbarnett · June 12, 2026 · 2 min
• "Estimating No-CoT Task-Completion Time Horizons of Frontier AI Models" by Anders Cairns Woodruff, Francis Rhys Ward, Dewi Gould, Rauno Arike, Jason R Brown, Jo Jiao, wlanderson, ariana_azarbal, harrymayne, Patrick Leask · June 11, 2026 · 10 min
• "Even “illegible” Mythos reasoning traces seem pretty legible" by faul_sname · June 11, 2026 · 8 min
• "Sequent: scale and automation for higher confidence in alignment" by Geoffrey Irving, Alex HT, Jesse Hoogland, Daniel Murfet, Jacob Pfau, Marco Cozzi, Stan van Wingerden · June 10, 2026 · 23 min
• "The Machines Lack Honour" by Raymond Douglas · June 10, 2026 · 20 min
• "My favorite depiction of utopia" by Caleb Biddulph · June 4, 2026 · 58 min