I canceled my OpenAI account. And here's why.

Disclaimer. Why this text is not "just drama."

I didn't cancel my OpenAI account out of frustration with a feature, a prompt, or a bill. It's not about complaining about some subscription, but about a development that goes much deeper: who actually controls this company today, what political alliances its leadership is forming, how its mission has shifted step by step - and where the journey is clearly intended to go economically.

It's important to me to separate this clearly: I'm not criticizing the technical performance of the products themselves, but the direction of the system that I help fund with every monthly payment. This text is not an impulsive rant, nor is it an attempt to persuade anyone to "boycott" anything. It is an organized collection of concrete points where I realized: my values and OpenAI's strategic direction have diverged so far that it would no longer be honest to continue paying silently.

In short: This is not a "cancel post" for clicks, but the documented rationale for a personal consequence - why I consciously decided to no longer financially support this company.

---

2. The moment I quit

The moment I quit

For a long time, OpenAI was "the" AI company for me: an early adopter, technically impressive, with a history that at least initially sounded idealistic - research for the common good, safety, no blind obsession with profit. I bought into the narrative that this was creating a counter-model to traditional Big Tech corporations: a lab that needed money but was primarily committed to a greater responsibility, not just the stock price.

I was a paying customer, went through GPT subscriptions, built workflows, produced content, developed strategies - OpenAI was deeply integrated into my daily life. Many of my processes ran "by default" on this infrastructure: from brainstorming texts and concepts to research, first drafts, and small automations. It would be a lie to say it was easy for me; on the contrary, precisely because the tools were so well integrated into my routines, I initially tended to push aside rather than actively question the developments in the background.

At some point, however, it became impossible to separate this usage from what was happening in the background: large political donations from the executive suite, mission statements that suddenly sounded different, and very concrete preparations for an IPO in the trillion-dollar range. The more pieces of the puzzle came together, the less I could convince myself that these were just "isolated incidents" or PR misunderstandings. It felt increasingly absurd to work with these tools during the day and read articles in the evening about precisely the developments I fundamentally disagreed with.

The point at which I canceled was therefore not a single scandal, but a tipping point: the feeling that a pattern had solidified. The political alignment at the top, the whitewashed mission, and the aggressive commercialization combined to form a picture that no longer aligned with what I wanted to spend my money on. At that moment, the logical consequence was to no longer be part of this model - even if it meant breaking cherished habits and finding other tools.

---

3. $25 Million for Trump - from the Executive Suite

Let's start with the most obvious and tangible point: money.

In autumn 2025, Greg Brockman, co-founder and president of OpenAI, personally donated $25 million to the pro-Trump Super PAC "MAGA Inc." According to FEC filings and consistent media reports, this was the largest single donation to that PAC during the period and accounted for almost a quarter of its total $102 million haul.

What's exactly going on here?

• It's not a corporate donation from "OpenAI Inc.," but a "private" donation from the president and co-founder.
• In effect, Brockman is one of the largest tech donors in Trump's circle.
• The donation goes to a Super PAC that politically supports Trump and, among other things, backs conservative candidates in midterm and special congressional elections.

It would be legally incorrect to say: "OpenAI donates $25 million to Trump." The correct statement is: "The president and co-founder of OpenAI gives $25 million to Trump's Super PAC."

Why is this a reason for me to cancel?

Because this isn't just any employee; it's the person at the very top who shapes the company - strategically, culturally, communicatively. When the president of one of the world's most powerful AI companies supports Trump with $25 million, that's a clear political statement.

Another detail that makes this even more bitter: Brockman's donation comes at a time when Trump-aligned circles are actively trying to block or dilute stricter AI regulations at the federal and state levels - something that is directly in the economic interest of companies like OpenAI. This isn't about abstract ideology; it's about concrete political ROI.

For me personally, this means: every euro I pay to OpenAI doesn't directly go into Trump's campaign - but I am funding the power base and ecosystem in which such decisions are made. That no longer feels neutral.

---

4. When a mission gets quieter: What OpenAI removed from its self-description

The second major block is less striking but at least as important: the mission.

In earlier years, OpenAI's mission was described in official documents as aiming to develop artificial general intelligence that "safely benefits humanity" and that it would "not be constrained by a need to generate financial return." These formulations are documented - including in older IRS forms (Form 990) for its non-profit structure.

In the current version, things look different:

• The word "safely" has been removed from the mission.
• The explicit commitment not to be constrained by a need to generate financial return has also been removed.
• The mission is reduced to a more general phrasing: essentially "ensure that artificial general intelligence benefits all of humanity" - without the added detail of how exactly and under what priorities (safety vs. return).

This isn't a meme, but a documented, analyzed process:

• Authors and analysts have explicitly pointed out that in its 2024 mission, OpenAI removed the word "safely" and the renunciation of profit orientation from its mission statement.
• In parallel, it was reported, among other things, that internal "Mission Alignment" teams were dissolved, while the corporate structure was adjusted towards classic, profit-oriented holdings.

Why is this relevant?

Because mission statements are not accidental. When you remove "safely" and "not profit-driven" from the official self-description of an AI corporation, that is a statement - even if you continue to speak of "responsibility" externally.

For me, this means: what once distinguished OpenAI from Big Tech corporations - the claim to be committed not primarily to shareholder value but to a higher mission - has been formally diluted. And precisely this point, this "we are not driven by profit," was central to why I initially placed my trust in the company.

---

5. From Lab to IPO Candidate: How OpenAI Moves Towards Wall Street

The third major reason is the path towards an IPO - not because an IPO is inherently bad, but because here it perfectly aligns with the other developments.

Several reports describe that OpenAI is specifically working on an IPO, which is expected to take place no earlier than 2026, more likely in 2026/2027 - with a target valuation of up to $1 trillion. Reuters and other media report that OpenAI is already in talks with US stock market regulators and banks and is considering raising at least $60 billion through the IPO, possibly significantly more.

In parallel, the corporate structure has been reorganized:

• OpenAI has established a Public-Benefit Corporation (OpenAI Group PBC), which acts as its for-profit arm.
• A new OpenAI Foundation (Non-Profit) holds about 26% of the shares and has certain control rights, while Microsoft owns about 27%.
• This structure is explicitly described as a way to maintain "mission oversight" while opening a clear path to public capital markets.

At the same time, estimates and analyses circulate suggesting that OpenAI aims for revenues in the range of $200 billion per year by 2030 - a level achieved by only a very few tech giants.

What does this mean in practice?

• A company that previously emphasized not being driven by the need for profit is now preparing for one of the largest tech IPOs in history.
• Large investors such as SoftBank, MGX, Thrive Capital, and others have already invested massive capital and stand to make gigantic potential profits from an IPO.
• Microsoft's stake of around 27% is still central after the restructuring - with a potential trillion-dollar valuation, we are talking about a theoretical value in the hundreds of billions for this stake alone.

For me, this is not just a financial detail, but a cultural signal: the closer a company gets to an IPO, the more it eventually has to meet the expectations of the capital market - a growth story, profit, scaling, margins. The mission "to benefit humanity" will inevitably stand alongside the stock price from this point on. Often, the stock price wins.

---

6. The Quiet Shift: From "Open" to "Closed"

Even without any headlines, you notice the development if you follow the product and the company for a while.

• OpenAI originally started with a strong open-source narrative, collaboration with research, publications, and a lab mindset.
• With the success of ChatGPT and massive investments, the model has increasingly shifted towards "Closed Source," proprietary APIs, and exclusive partnerships with large corporations (foremost Microsoft).
• At the same time, alternatives have grown - open models, community projects, foundation models published under permissive licenses.'

In combination with the new mission, the proximity to Trump at the top, and the IPO plans, a picture emerges for me: OpenAI is no longer the research-driven "public good" it once claimed to be, but a classic tech corporation with maximum economic leverage - only with significantly more power than a normal app company.

---

7. Trust cannot be arbitrarily divided

One could now say: "But the product is good, why do you care about a president's politics or the wording in an IRS mission?"

My problem is: generative AI is not just a tool, but infrastructure for information, communication, creativity - in short: something that penetrates deeper and deeper into our daily lives and our decisions.

If I, regarding such an infrastructure:

• trust it to see my data, process my texts, know my thoughts,
• at the same time see that at the top there is a person who supports Trump with $25 million,
• and the company rephrases its mission so that safety and independence from profit are no longer explicitly central,
• while it is preparing for a gigantic IPO,

then for me it's no longer a "neutral cloud service," but a political-economic actor with a clear direction.

I have no problem with companies wanting to be profitable. I have a problem when a company that shapes the flow of information in society is simultaneously moving towards maximum capital market pressure and proximity to Trump - and in the fine print, becomes quieter when it comes to "safely" and "not profit-driven."

---

8. Why canceling is the logical step for me

Canceling doesn't change the world or stop an IPO. But it changes one thing: where my money goes - and what I indirectly help legitimize with it.

By canceling, I'm essentially saying four things:

1. I no longer accept that someone in the top leadership of an AI giant supports Trump with $25 million - and that I simultaneously act as if this has nothing to do with the company I pay for.
2. I acknowledge that OpenAI has rephrased its mission so that safety and independence from profit expectations are no longer explicitly enshrined - and I find this unacceptable for an AI company of this size.
3. I see that OpenAI is preparing for an IPO with a trillion-dollar valuation and thus fully enters the logic of capital markets - which inevitably shifts priorities.
4. I know that there are alternatives - technically and organizationally - and that my subscription is not a law of nature.'

This is not a call for everyone to cancel OpenAI now. It is an explanation of a personal decision for consistency: I don't want to talk about politics, power, and ethics on the one hand - and on the other hand, quietly pour money every month into a system whose direction I find problematic.

---

9. What I do instead

Canceling is only half the story. The other half is: Where do I go? And how do I put my principles into practice?

I now orient myself on three clear guidelines - and yes, that also means testing new tools, changing habits, and sometimes making compromises:

Transparency: I prefer providers who disclose how their models are trained, what data is used, and how they handle security - and whose mission doesn't quietly delete important terms like "safely." No black box where I can only guess what's going on in the background.

Structure: The structure of a company has become more important to me - who holds the shares, whether there's a genuine non-profit component, or if it's a classic "growth story for an IPO" with Microsoft-like dependencies. I look for models that are not primarily beholden to the next funding round.

Promote Alternatives: Where possible, I want to support open models, smaller providers, and community projects that create real competition - instead of pushing everything into a single, highly concentrated center of power. This explicitly includes local models: tools like Ollama, LM Studio, or self-hosted instances of Llama/Mistral that I run on my computer. No cloud, no API keys, no monthly bills - just models that work offline and give me full control.

A central theme here is avoiding vendor lock-in: After years where OpenAI dictated my workflows, I'm now building on standardized interfaces (e.g., OpenAI-compatible APIs from alternatives like Grok, Anthropic, or local hosts) and open-source stacks. This allows me to switch quickly without rewriting everything. Examples from my rotation:

• Anthropic (Claude) for complex texts and reasoning, with a clearer security focus.
• Mistral, Kimi for creative and humorous tasks, less "politically charged."
• Local models (e.g., Llama 3.1 via Ollama) for sensitive or recurring tasks - 100% data privacy compliant, zero vendor dependency.
• Additions like Perplexity for research, Hugging Face for specialized models.

This isn't always as comfortable as the big, convenient player with its seamless integration. And, yes, even with other providers (Grok, Anthropic/Claude), you should look twice - especially with Claude, which was reportedly (contrary to Anthropic's consent, see sources 16 and 17) used in a US military operation to apprehend Nicolás Maduro in Venezuela. Sometimes I have to adapt prompts, accept latency, or combine multiple tools. But it feels more honest - and more robust in the long run. Less dependency means more freedom, and that's invaluable in an AI world that is changing at breakneck speed.

---

10. Conclusion: One less subscription, a little more control

In the end, my cancellation is not a heroic act, but a very mundane one: I closed an account and stopped a payment. It wasn't a dramatic event, no social media statement with a screenshot of the cancellation email, no call to the mob. Simply: deactivate subscription, close browser tab, get back to work.

But precisely in an environment where an OpenAI president supports Trump with $25 million, where mission statements lose precise formulations like "safely" and "unconstrained by a need to generate financial return," and where the stage is set for a trillion-dollar IPO, this is my way of drawing the consequence from these facts. These points - large political donations from the leadership, a subtly adapted self-description, the march towards Wall Street - are not isolated. Together, they paint a clear picture of shifting priorities that I eventually could no longer ignore.

I don't want my money - however small the amount on a global scale - to be part of this system. It's not about "bringing down" OpenAI or portraying myself as a martyr. It's about personal coherence: If I talk about ethics, responsibility, and power in tech on the one hand, and on the other hand, let 20 euros flow every month in exactly the direction I find problematic, then I lose credibility - first with myself.

That's why I canceled my OpenAI account. And that's exactly why I stand by it. It doesn't feel like a loss, but like a small, yet genuine act of self-determination: Less dependence on a monopolist, more choice of tools, more conscious questioning of what I use and support. In an era where AI is becoming more and more commonplace, that is perhaps the most important gain - a little more control over what controls me.

---

Sources:

1. Greg Brockman's $25M donation to Trump's Super PAC, classification and quotes https://www.wired.com/story/openai-president-greg-brockman-political-donations-trump-humanity/[wired]' German summary / quotes ("Team Humanity") https://t3n.de/news/openai-chef-millionenspende-trump-unpolitisch-1729498/[t3n]'
2. Details on $25M to MAGA Inc., share of war chest, FEC context https://www.sfgate.com/tech/article/brockman-openai-top-trump-donor-21273419.php[sfgate]' Brief summary in financial press https://finance.yahoo.com/news/openai-exec-becomes-top-trump-230342268.html[finance.yahoo]' Further summary https://news.aibase.com/news/24965[news.aibase]'
3. **Background on Trump's Super PAC and record fundraising **https://www.brennancenter.org/our-work/analysis-opinion/pro-trump-super-pac-raises-record-breaking-305-million[brennancenter]'
4. **Short news on "OpenAI President drops $25M on Trump Super PAC" **https://www.techbuzz.ai/articles/openai-president-drops-25m-on-trump-super-pac[techbuzz]'
5. **Summary of the donation in social posts (as additional source) **WIRED-Post: https://www.linkedin.com/posts/wired_openais-president-gave-millions-to-trump-activity-7427805539449090048-blNk[linkedin]' Instagram-Post: https://www.instagram.com/p/DUvpG6lDBSY/[instagram]'
6. Analysis: "OpenAI has deleted the word 'safely' from its mission" (structure vs. shareholders) https://theconversation.com/openai-has-deleted-the-word-safely-from-its-mission-and-its-new-structure-is-a-test-for-whether-ai-serves-society-or-shareholders-224785[theconversation]' Syndicated version (The Week): https://www.theweek.in/wire-updates/international/2026/02/15/openai-has-deleted-the-word-safely-from-its-mission.html[theweek]'
7. **Report: "OpenAI has deleted the word 'safely' from its mission - Yahoo Finance" **https://finance.yahoo.com/news/openai-deleted-word-safely-mission-132143762.html[finance.yahoo]'
8. **Further analysis of mission change (Blog) **https://ai-minor.com/blog/en/2026-02-14-1771042834237-openai_has_deleted_the_word__safely__from_its_miss/[ai-minor]'
9. **Discussion of mission change (Community context) **https://www.reddit.com/r/ArtificialInteligence/comments/1r4kkyc/openai_dropped_word_safely_from_its_mission_meta/[reddit]'
10. **OpenAI's own structure/governance contribution ("Why our structure must evolve") **https://openai.com/index/why-our-structure-must-evolve-to-advance-our-mission/[openai]'
11. **Background on OpenAI, history & structure **https://en.wikipedia.org/wiki/OpenAI[en.wikipedia]'
12. **Reuters Exclusive Report: "OpenAI lays groundwork for juggernaut IPO at up to $1 trillion" **https://www.reuters.com/business/openai-lays-groundwork-juggernaut-ipo-up-1-trillion-valuation-2025-10-29/[reuters]' ***Reflected on Yahoo Finance: ***https://finance.yahoo.com/news/exclusive-openai-lays-groundwork-juggernaut-232125990.html[finance.yahoo]'
13. **SiliconANGLE: "OpenAI reportedly planning 2026 - 2027 IPO at valuation of up to $1T" **https://siliconangle.com/2025/10/29/openai-reportedly-planning-2026-2027-ipo-valuation-1t/[siliconangle]'
14. **Further financial/stock market reports on the planned IPO **https://finance.yahoo.com/news/openai-eyes-massive-1t-ipo-175600009.html[finance.yahoo]' https://www.bloomberg.com/news/articles/2025-10-29/openai-could-target-1-trillion-value-in-ipo-reuters-says[bloomberg]' ***(Reflected in financial portals, e.g., Investing.com) ***https://www.investing.com/news/stock-market-news/openai-seen-preparing-for-2027-ipo-at-1-trln-valuation-reuters-4317757[investing]'
15. **Summarized German-language reports on Brockman's donation **https://t3n.de/news/openai-chef-millionenspende-trump-unpolitisch-1729498/[t3n]'
16. Small Wars Journal / Wall Street Journal (Feb 16, 2026): "Pentagon Used Anthropic's Claude in Maduro Venezuela Raid" https://smallwarsjournal.com/2026/02/17/ai-enabled-decapitation-strike-maduro-raid/
17. 'NDTV / Wall Street Journal (Feb 13, 2026): "US Military Used Claude AI In Venezuela Operation To Capture Nicolas Maduro" https://www.ndtv.com/world-news/us-military-used-claude-ai-in-venezuela-operation-to-capture-nicolas-maduro-report-11003847