Bloomberg Businessweek – Episode Summary

Episode: Trump Orders US Agencies to Drop Anthropic After Pentagon Feud
Date: February 27, 2026
Hosts: Carol Massar and Tim Stenovec

Overview

This episode centers on two major business and tech stories:

1. President Trump's order for federal agencies to stop using Anthropic’s AI products, following a high-stakes standoff between Anthropic and the Pentagon over the use of AI in military applications.
2. Broader discussions on tech and economy, including an in-depth interview with Duolingo’s CEO on their AI and growth strategy, and an exploration of a major housing partnership aimed at addressing affordability.

Key Segment 1: President Trump & Anthropic-Pentagon Standoff (02:12–09:44; returned to topic at 40:44–41:37)

Main Points

• Trump’s Order: President Trump directed all US government agencies to phase out use of Anthropic’s AI products over six months, especially in the Pentagon, citing concerns over adhering to the Constitution versus AI vendor terms. Trump’s post on Truth Social stated:

  "The left wing nut jobs at Anthropic have made a disastrous mistake trying to strong arm the Department of War and force them to obey their terms of service instead of our Constitution. Their selfishness is putting American lives at risk." — President Trump [02:18]

• Nature of the Dispute:

  • The Pentagon wants flexible use of AI, potentially including autonomous systems; Anthropic refuses to allow its technology to be used for fully autonomous lethal weaponry.
  • Anthropic has a significant role in defense tech stacks (e.g., Palantir’s Maven Smart), making their removal complex and risky for continuity.

• Negotiations and Deadlines:

  • Pentagon and Anthropic reportedly still in last-minute negotiations per an Axios report.
  • Defense Secretary Pete Hegseth threatened to declare Anthropic a "supply chain risk" if a deal was not reached by the 5:01 PM deadline.

• Stakeholder Perspectives:

  • Some officials characterize Trump’s move as politicized, not just about security (per Senate Intelligence Committee input).
  • Katrina Manson, Bloomberg reporter and author on military AI, emphasizes how embedded Anthropic is and the underlying military hesitancy about the technology’s readiness.
  • Both Anthropic and some military leaders stress concerns over AI autonomy, not purely political or commercial issues:

    “They didn’t want their tools being involved in autonomous warfare because they’re not ready.” — Katrina Manson [05:44]
    “The technology isn’t ready yet.” — Jack Shanahan, former Project Maven leader, paraphrased by Manson [06:33]

• Complications of Transition:

  • Six-month phase-out would be operationally complex.
  • Anthropic has offered to support transition and help improve system reliability, but Pentagon has so far refused R&D collaboration.
  • The fundamental disagreement reportedly centers on whether “human judgment” will always apply to use of force.

• Notable Quote:

  “They are saying 'we want to help you use this wild technology in a reliable way and help us help you.’” — Katrina Manson on Anthropic’s R&D proposal [09:25]

Notable Timestamps

• 02:12 — Trump’s directive and context
• 03:50 — Katrina Manson explains Anthropic’s entrenchment in defense and transition issues
• 05:29 — National security implications and readiness concerns
• 07:38 — The performative nature of the standoff and the deadline’s meaning
• 09:19 — Anthropic’s offer to collaborate and Pentagon’s response
• 40:44 — Recap on Trump’s post and restatement of Anthropic’s red lines

Key Segment 2: Compass & Rocket Companies on Housing Affordability (13:36–27:13)

Main Points

• Strategic Partnership Announced:

  • Compass (largest real estate brokerage) and Rocket (largest mortgage lender) unveil a three-year alliance to improve home affordability and inventory.
  • Aim to bring 500,000 to 1 million previously unavailable home listings to market, focusing on private exclusives and “coming soon” listings.

• Affordability Focus:

  • The initiative seeks to lower barriers for sellers and provide mortgage incentives (up to $6,000 closing coverage, rate discounts).
  • Target both sides of the market — unlocking inventory for buyers and creating new tools for sellers.

• Changing the System:

  • Proposal to eliminate “days on market” and “price drop history” visibility for sellers, mirroring practices in markets like Australia.
  • The intent is to reduce sellers’ hesitancy and increase inventory, though some hosts question if reduced transparency could mislead buyers or contribute to price inflation.

• Industry Reform Requests:

  • Robert Reffkin (Compass CEO) urges abolition of the National Association of Realtors’ “clear cooperation” rule, which mandates MLS listings, arguing for seller autonomy.

• Notable Quotes:

  • “There should not be these artificial barriers... We are taking [negative insights] off the listing.” — Robert Reffkin [14:09]
  • “It’s really important, I think, to acknowledge that and to diversify the way you qualify folks.” — Varun Krishna, Rocket CEO, on credit and income evaluations [17:35]
  • “We believe if you, the buyer, what you would you need is you need more inventory. So to help you, the buyer, the best thing we can do is just let the seller sell however they want.” — Reffkin [19:42]
  • “There should be no fear to list. There should be no downside to list. There should be no barrier... If you take away all those barriers, you will have more inventory.” — Reffkin [22:20]

Notable Timestamps

• 13:36 — Introduction to partnership
• 14:33 — Reffkin outlines partnership impact for sellers and buyers
• 15:45 — Krishna details inventory unlock and mortgage incentives
• 17:35 — Discussion on expanding mortgage qualification criteria
• 22:20 — Emphasis on removing barriers for sellers

Key Segment 3: Duolingo’s CEO on AI, Growth & Investor Concerns (30:29–37:58)

Main Points

• AI Strategy:

  • CEO Luis von Ahn emphasizes Duolingo’s focus on leveraging AI for superior teaching, aspiring to an app as effective as a human tutor and as engaging as a game.
  • Duolingo holds vast data on user learning, giving it an edge in AI training.

• Competitive Position:

  • Little concern over new AI entrants in language learning; company has 85% market share and billions of user interactions for model improvement.
  • Points to the distinctive value in fostering user streaks and learning habits, not just content delivery.

• Growth and Monetization:

  • Short-term profit margins will be sacrificed for long-term user growth (goal: double active daily users to 100 million by 2028).
  • Expansion to new disciplines (math, chess, music); currently ~15% of active users engage outside language learning.

• Investor Misunderstandings:

  • Stock is down due to a forecast for slower short-term growth, but leadership underscores a long-view strategy.
  • Shift this year to focus on “growing the pie” (total active users) rather than monetizing the existing base.

• Notable Quotes:

  • “Because of AI, we’re going to be able to teach significantly better... our sense is that within the next few years, we’re going to have an app that teaches as well as a one-on-one tutor, but also is as fun as a mobile game.” — Luis von Ahn [31:26]
  • “The hardest thing about learning something by yourself is staying motivated... we motivate people to continue doing it. And that’s hard to get right.” — Luis von Ahn [33:09]
  • “One of the main advantages that we have certainly against new entrants is scale. There’s no other education app that has the scale that we have.” — Luis von Ahn [37:41]

Notable Timestamps

• 31:26 — Vision for AI in EdTech
• 32:12 — Duolingo’s competitive advantage rooted in scale and data
• 33:09 — Importance of habit formation and motivation
• 34:13 — On investor skepticism and long-term strategy
• 36:31 — Discussion of new subject areas and future growth

Key Segment 4: Market Analysis — AI, Credit, and Employment Risks (42:11–49:46)

Main Points

• Financial Markets & AI Risk:

  • Uncertainty reigns in markets due to multiple factors: risks in private credit, rising loan delinquencies, and reliance on private funding for tech giants like Anthropic/OpenAI.
  • AI spending is crucial to tech-sector growth, but layoffs (e.g., Square’s 40% workforce reduction) stoke fears about automation’s threat to white-collar jobs.

• Broader Economic Implications:

  • The panel reflects on historical tech disruption: while previous innovations eventually created jobs, current AI trends aim directly at eliminating intermediary (white-collar) work, raising deeper economic and social concerns.
  • Lack of transparency in the financial deals sustaining AI giants sparks additional market anxiety.

• Notable Quotes:

  • “If all these companies can do more with half the number of employees, where do those employees go and then what does it mean for their spending power?” — Tim Stenovec [45:13]
  • “This specific technology is designed to make things more efficient… which is to eliminate the need for you and I to be intermediaries and that’s white collar jobs… this could actually go right after those people that we need to be spending, which is scary.” — David Gura [46:26]

Notable Timestamps

• 42:11 — Market trends, credit worries, the AI funding cycle
• 44:48 — Concern about job destruction in the age of AI
• 45:44 — How AI efficiency affects labor and consumer spending
• 46:12 — Comparison to past technological disruptions

Memorable Moments

• Trump’s no-holds-barred attack on Anthropic as “left wing nut jobs” accused of putting “American lives at risk” [02:18]
• Katrina Manson’s macro perspective: “It’s very significant that the President’s weighed in and called this directive… unraveling Anthropic over the course of six months will be extremely complicated… They are used in combat operations on Maven Smart system which is a Palantir system.” [03:50]
• Concise summary of what’s at stake for the Pentagon: “The technology isn’t ready yet.” — Jack Shanahan, paraphrased [06:33]
• Robert Reffkin’s real estate paradigm-shift: “You should be able to list however you want... if we do that, there’ll be more inventory.” [22:20]
• Luis von Ahn on AI and motivation: “The hardest thing about learning something by yourself is staying motivated… that’s what we do differently.” [33:09]

Conclusion

This episode delivers a front-row seat to current business and tech turmoil:

• The collision of political directive and private AI development, and its knock-on effects for national security and federal tech policy
• A transformative alliance in real estate aimed at remedying America’s housing crisis
• Candid CEO perspectives on AI’s power and uncertainty, both in edtech and beyond
• Persistent worries (and some optimism) regarding the broader economic effects of AI-driven change

For anyone tracking the intersections of Silicon Valley, Washington, housing, or financial markets, these conversations are both urgent and instructive.