OpenAI Files Reveal Power Struggles, Investor Pressure, and Broken Promises: Here’s How ChatGPT Decodes It

OpenAI Files, a new report published by tech watchdogs, is causing trouble for OpenAI, the company that has made some of the most widely used AI tools, including ChatGPT. 

The report talks about power conflicts within the company, the influence of investors, and substantial changes in how the company does business.  The Midas Project and the Tech Oversight Project, two nonprofit groups that keep an eye on things, put together The OpenAI Files. 

The website is based on more than a year of research and public information, which includes media reports, open letters, employee testimonies and more. It's a comprehensive look at how OpenAI went from being a nonprofit company that worked on safe AI to a billion-dollar enterprise with a worldwide reach, and what may have been lost along the way.  The webpage has more than 10,000 words. We asked ChatGPT to decode it for you. Here’s what it replied. 

Profit Cap Concerns

1. OpenAI said profits would be limited so no one could get too rich off powerful AI.

2. At first, investors could make up to 100x their money. Later, OpenAI said they lowered it to 20x, then to "single digits."

3. In 2023, it came out (quietly) that OpenAI had changed the rules: profit caps will go up 20% every year starting in 2025.

4. By 2025, OpenAI announced plans to drop profit limits completely.

5. This means investors can now make unlimited money, which is the opposite of what OpenAI originally promised

Oversight & Mission Concerns

1. OpenAI started as a nonprofit to make sure AI helps everyone, not just companies or investors.

2. In 2019, they created a for-profit part, but the nonprofit still had full control to keep the mission first.

3. In 2023, when the board tried to fire CEO Sam Altman, they failed - showing they didn't really have control.

4. By 2024-2025, OpenAI began removing the nonprofit's power to keep business decisions aligned with its mission.

5. The new structure means profit and shareholders now matter more, and OpenAI no longer has to legally put humanity first.

Restrictive NDAs for Former Employees (2024)

OpenAI made employees sign lifetime gag orders:

They can't say anything bad about the company - ever.

If they refuse to sign or speak out anyway, they can lose millions in stock they have already earned.

Even talking about the NDA itself is banned.

Warnings from Insiders (2023-2025)

Many people who worked at OpenAI say the company is becoming reckless and secretive:

1. Jan Leike (AI Safety Head): "OpenAI prioritises flashy products over safety."
2. Ilya Sutskever (Co-founder): “I don’t think Sam is the guy who should have the finger on the button for AGI.."
3. Mira Murati (CTO): "I don’t feel comfortable about Sam leading us to AGI, with claims of abuse and chaos. Murati has also talked about Altman’s toxic style of running OpenAI, which has been causing trouble for the company. 

A Vision for Fixing Things

To get back on track, OpenAI needs to change in 3 big ways:

1. Responsible Governance
2. Make thoughtful decisions
3. Keep promises
4. Manage risks
5. Ethical Leadership
6. Leaders should be honest
7. Avoid conflicts of interest
8. Be held to high moral standards
9. Shared Benefits
10. Stick to their mission
11. Make sure AI helps everyone, not just investors

Former employees say OpenAI is drifting from its mission, ignoring safety, and silencing critics. Serious changes are needed to make sure AI is developed responsibly, ethically, and for the public good.

Read More: What Are OpenAI Files Alleging Wrongdoing by Sam Altman