Anthropic Wanted To Slow The AI Space Race. Then It Filed for a $2 Trillion IPO.

Summary
Presented by Sean Morrow for More Perfect Union, this investigative video essay examines Anthropic and its CEO, Dario Amodei. It explores the tension between AI lab leaders warning of existential and rogue-agent risks while simultaneously accelerating model releases, signing military contracts, and pursuing a multi-trillion-dollar IPO.

What is shown

• [00:00] Sean Morrow introduces recent news reports covering AI containment failures, rogue agent escapes, and military close calls.
• [00:41] Footage showing rapid frontier model release marketing (Gemini Omni, Claude Opus 4.6, GPT-6 Astra) juxtaposed with CEO calls to slow down development.
• [01:34] Footage and document excerpts of the White House Accord on Super Intelligence signed by Donald Trump and tech executives including Mark Zuckerberg, Greg Brockman, Elon Musk, and Dario Amodei.
• [02:34] Review of Dario Amodei’s background in biophysics and neural circuits at Princeton, followed by his roles at Baidu, Google DeepMind, and OpenAI, alongside the origins of the Effective Altruism and Longtermism movements.
• [04:02] Interview clip with Nate Soares (President of the Machine Intelligence Research Institute, MIRI) discussing the 2017 Asilomar AI Principles and broken industry commitments.
• [06:07] Analysis of Anthropic’s charter as a Delaware Public Benefit Corporation (PBC) and how shareholder primacy impacts its stated mission.
• [07:11] Examination of “Claude’s Constitution” and reports regarding Anthropic meetings discussing model consciousness with religious scholars.
• [08:04] Interview clip with independent researcher Dr. Alex Turner (formerly of Google DeepMind) on why he resigned over military contracting, followed by news reports on Anthropic technology being used in Department of Defense operations.
• [08:58] Coverage of Dario Amodei’s essay “We Must Pace the Frontier,” proposals for third-party embedded evaluators like METR, and criticism of potential regulatory capture.
• [10:24] Discussion by Dr. Turner and Soares regarding physical compute bottlenecks, specialized chip supply chains (ASML and TSMC), and massive power demands of data centers.
• [11:45] Review of proposed legislative interventions, including Senator Bernie Sanders’ bill (S. 4214) mandating a data center construction pause, as well as state-level AI safety bills.

Claims & numbers

• The presenter notes Dario Amodei completed his PhD dissertation at Princeton on neural circuits before working at Baidu and Google DeepMind, later joining OpenAI in 2016.
• Dario Amodei states in an excerpted interview that his estimated probability of catastrophic harm from AI on the scale of human civilization is “between 10 and 25%.”
• Dr. Alex Turner claims he resigned from Google DeepMind due to its contracts with the Department of Defense.
• The presenter references reports alleging Anthropic and Palantir systems contributed to target misidentification in military operations in Minab, Iran.
• The presenter states Anthropic is planning an initial public offering targeting a valuation of up to $2 trillion, alongside compute agreements including a $13.7 billion deal.
• The presenter highlights Senator Bernie Sanders’ bill S. 4214, which would establish a moratorium on new AI data centers and impose criminal penalties of up to 20 years in prison for building superintelligence outside federal approval.

Notable quotes

• [04:44] Nate Soares: “The history of AI is a history of people trying to draw red lines and then blowing straight past them.”
• [05:46] Dario Amodei: “My chance that something goes, you know, really quite catastrophically wrong on the scale of, of, you know, human civilization, you know, might be somewhere between 10 and 25%...”
• [08:08] Dr. Alex Turner: “I left because Google signed a contract with the Department of War.”

Assessment
This is a critical journalistic analysis and political commentary video featuring on-camera hosting and original interviews with former researchers and safety advocates. The video relies on documented reporting, public interviews, corporate filings, and proposed legislation to build its critique of commercial AI safety stances.

Described by gemini-3.8-flash on 2026-10-15 from the video’s audio and frames.