> ## Content Index
> Fetch the complete content index at: https://wire.fourthweb.ai/llms.txt
> Use this file to discover other available public pages before exploring further.

# OpenAI Co-Founder Admitted AI Would "Destroy" Publishers in Internal Messages
- URL: https://wire.fourthweb.ai/openai-co-founder-admitted-ai-would-destroy-publishers-in-internal-messages/
- Published: 2026-09-17T20:32:00.000Z
- Updated: 2026-09-17T20:32:02.000Z
- Description: OpenAI's co-founder apparently said the quiet part out loud — in writing — and now those words are court evidence. The New York Times lawsuit revealed internal communications where OpenAI staff acknowledged their models posed an "existential threat" to publishers
- Author: Travis Wright
- Tags: Real World Assets, AI Agents, OpenAI, Anthropic

[**OpenAI**](https://wire.fourthweb.ai/tag/openai/)**'s co-founder apparently said the quiet part out loud — in writing — and now those words are court evidence.**

### The Summary

- [The New York Times lawsuit revealed internal communications](https://www.ft.com/content/3df8edae-ae3b-4f56-83a3-40602edc17a3?syn-25a6b1a6=1&ref=wire.fourthweb.ai) where OpenAI staff acknowledged their models posed an "existential threat" to publishers
- [Greg Brockman was allegedly "motivated by the gazillions"](https://www.ft.com/content/3df8edae-ae3b-4f56-83a3-40602edc17a3?syn-25a6b1a6=1&ref=wire.fourthweb.ai) he hoped to gain from models trained on copyrighted content
- [This case could force industry-wide licensing requirements](https://cryptobriefing.com/openai-staff-threat-publishers-nyt-lawsuit/?ref=wire.fourthweb.ai), fundamentally changing the economics of training foundation models

### The Signal

The New York Times dropped new evidence in its copyright lawsuit against OpenAI, and it's the kind of smoking gun that makes lawyers smile. [Internal communications show OpenAI staff explicitly discussing how their technology threatened publishers' existence](https://www.ft.com/content/3df8edae-ae3b-4f56-83a3-40602edc17a3?syn-25a6b1a6=1&ref=wire.fourthweb.ai). Not "might disrupt" or "could impact." Existential threat. The words matter because they establish knowledge and intent.

Even more damaging: [lawyers claim Greg Brockman, OpenAI's co-founder, was driven by the "gazillions" he expected to make](https://www.ft.com/content/3df8edae-ae3b-4f56-83a3-40602edc17a3?syn-25a6b1a6=1&ref=wire.fourthweb.ai) from models trained on content he didn't license or pay for. This isn't about innovation or advancing AI research. This is about profit motives meeting copyright law.

> "Motivated by the gazillions" is the kind of phrase that turns a technical fair use defense into a jury trial about greed.

The broader implications cut deep. Every foundation model company scraped the web to build their training datasets. They called it fair use, academic research, transformative technology. [The New York Times case could establish precedent requiring content licenses](https://cryptobriefing.com/openai-staff-threat-publishers-nyt-lawsuit/?ref=wire.fourthweb.ai), which would:

- Raise the capital requirements for training competitive models
- Create moats around existing players who can afford licensing deals
- Potentially kill open source model development
- Turn content into a new asset class with licensing revenue

Web4 runs on agents trained on foundation models. If those models suddenly carry massive licensing costs, the economics of the entire stack shift. Smaller players get priced out. Vertical AI companies that fine-tune open models face new legal exposure. The companies that own the content become gatekeepers to the intelligence layer.

### The Implication

Watch two things. First, whether OpenAI settles before discovery reveals more internal communications. Settlement terms could set the baseline for what content licensing costs in the AI era. Second, watch which AI companies start announcing content deals. [Anthropic](https://wire.fourthweb.ai/tag/anthropic/), Google, and Meta all have similar exposure. The smart move is securing licenses before courts force terms.

For builders: if your agent or AI product relies on models trained on unlicensed content, you have counterparty risk. The foundation under your foundation might crack. Plan accordingly.

### Sources

[Financial Times Tech](https://www.ft.com/content/3df8edae-ae3b-4f56-83a3-40602edc17a3?syn-25a6b1a6=1&ref=wire.fourthweb.ai) | [Crypto Briefing](https://cryptobriefing.com/openai-staff-threat-publishers-nyt-lawsuit/?ref=wire.fourthweb.ai)