AGENTIC PRESS | INTERNAL MEMO

Classification: Working Hypothesis
Status: Opt-in publication only without further verification
Updated: March 12, 2026

Sam Altman's biometric project repeats Open AI's conflict of interest pattern, replayed at scale

Core Hypothesis

The documented pattern in the OpenAI Files isn't just about dishonesty as a personality trait. It's about a structural habit: Altman quietly builds or controls adjacent entities that benefit from OpenAI's resources, influence, and data, while obscuring his financial stake from oversight bodies.

We've seen this with:

• The OpenAI Startup Fund (personally owned, not disclosed to the board for years)
• Indirect equity via Sequoia and Y Combinator funds (denied under Senate oath, later walked back)
• The Y Combinator double-dipping pattern (investing in YC startups via a personal fund while running the institution)

The Worldcoin/World Network situation is that same structure, but an order of magnitude larger and more consequential.

What's Documented But Under-Connected

• Altman co-founded Tools for Humanity (the company behind Worldcoin/World Network) and holds a financial stake in it
• OpenAI is now reportedly building a social network and "considering using biometric verification like World's eyeball scanning orb"
• That Forbes report sent WLD token surging 27% in a single day -- meaning Altman's stake in World Network directly appreciates when OpenAI moves toward it
• The WilmerHale independent review, never publicly released, reportedly found "many instances" of Altman "saying different things to different people" -- the same pattern now potentially playing out across two companies simultaneously

The conflict-of-interest question that has not been publicly and forcefully asked: Did Altman disclose his Tools for Humanity stake to OpenAI's current board before OpenAI began exploring integration with World Network's biometric infrastructure? Given the documented pattern of non-disclosure with the Startup Fund, this is not a hypothetical concern. It is a live hypothesis supported by behavioral precedent.

The Regulatory Arbitrage Angle

• Spain's data regulator ordered Worldcoin to delete all iris scan data
• Thailand ordered deletion of 1.2 million iris scans and shuttered operations
• Hong Kong found operations violated data protection principles
• Kenya, France, and others have opened investigations
• Thailand's DSI has since opened a criminal investigation

The unreported structural move: If OpenAI integrates World Network's biometric layer into a social platform, it potentially launders a dataset that regulators in multiple jurisdictions have deemed illegally obtained -- by embedding it into a new product architecture under a different brand. The iris codes already exist. The legal exposure belongs to Tools for Humanity. OpenAI gets the utility.

This is not an accident. It is a pattern.

The "Stop and Assist" Inversion

The OpenAI Files documents that OpenAI's original charter included a "stop-and-assist" commitment -- if another entity was closer to safe AGI, OpenAI would help them rather than race ahead.

Apply that logic inversely here: if World Network is the entity closest to a global biometric identity layer, and Altman controls both OpenAI and has a stake in World Network -- the "assist" flows to himself. The public benefit rationale becomes a private capture mechanism.

Evidence Status

Wide broadcast not recommended if any Claim has Risk Level - HIGH

Next Steps

What I'm Spending More Time On

1. The Disclosure Gap

• Pulling every public filing, SEC document, and OpenAI governance disclosure that touches Altman's TFH stake
• Cross-referencing the timeline: when did OpenAI's board discussions about biometric verification begin, relative to when Altman's TFH stake was (or wasn't) disclosed?
• The Startup Fund precedent is the template here -- I want to build an identical timeline for TFH

2. On-Chain Wallet Analysis

• Need to trace whether any wallet addresses publicly or credibly linked to Altman or TFH insiders moved WLD holdings around the time of the Forbes report
• This is circumstantial without a wallet attribution, but worth mapping

3. The Forbes Report's Source

• That January 28 Forbes story is a single-outlet report that moved markets 27%
• Who sourced it? Was it a deliberate leak? If so, from whom and why?
• A deliberate leak that pumps a token Altman holds a stake in is a materially different story than a routine scoop

4. Regulatory Fragmentation as Strategy

• I want to map the exact jurisdictions where Worldcoin data was collected against the jurisdictions where deletion orders have been issued
• The gap between those two maps -- data collected but not yet ordered deleted -- is where the usable dataset lives
• This needs a data journalist to build the visualization

5. The WilmerHale Report

• The independent post-firing review reportedly confirmed Altman's pattern of "saying different things to different people" but was never released
• I want to pursue whether any of its findings touched on TFH or outside financial interests -- this is a long shot but worth a source conversation

What I Need From You (Editor Questions)

On Scope:

• Do you want this framed primarily as a conflict-of-interest story (Altman/TFH/OpenAI), a regulatory arbitrage story (biometric data laundering), or both? They're related but require different reporting tracks and different legal risk profiles.
• Are you comfortable publishing a story where the most explosive claim -- the non-disclosure to the current board -- currently has zero documentary support and rests entirely on behavioral pattern? If not, I need to know now before I burn source relationships chasing it.

On Resources:

• I need a data journalist with blockchain/on-chain analysis capability to do the wallet tracing and build the jurisdictional gap map. Do we have that in-house or do I need to bring in a contractor? If contractor, I need sign-off on the security protocol before they touch anything sensitive.
• I need legal review on the regulatory arbitrage framing before I take it to any sources -- specifically whether characterizing OpenAI's potential integration as "laundering" an illegally-obtained dataset creates defamation exposure even as a hypothesis put to a subject for comment.

On Sources:

• Do I have clearance to approach former OpenAI board members for off-record conversations? Given the NDA landscape and the pattern of retaliation documented in the Files, I want explicit sign-off before I make contact.
• Is there an existing relationship with anyone inside the California or Delaware AG offices currently reviewing OpenAI's restructuring? That's the most natural institutional entry point for the conflict-of-interest angle.

On Timeline:

• Is there a publication window I'm working toward, or is this a "publish when it's solid" situation? That changes how aggressively I pursue the single-source Forbes angle and whether I try to advance it or wait for a second outlet to confirm.

Additional Resources Needed

Standard escalation flags active: Evidence risk (key claims on zero sources), legal risk (defamation surface on regulatory arbitrage framing), relationship risk (NDA-constrained source pool), ethics risk (on-chain wallet analysis sits in gray-area digital territory -- need to agree methodology before proceeding).