Skip to main content
Sam Altman, OpenAI CEO and Trump mega-donor, with other tech executives at a White House event [nytimes.com](https://www.nyti

Editorial illustration for OpenAI President, Trump Mega-Donor, Joins Tech Execs Courting Administration

OpenAI President's Trump Donation Sparks Tech-Politics Clash

OpenAI President, Trump Mega-Donor, Joins Tech Execs Courting Administration

Updated: 3 min read

OpenAI president Greg Brockman is cutting two very large checks. One, for millions, went to a super PAC targeting New York Assemblymember Alex Bores, who cosponsored a state AI regulation bill. The other, a mega-donation, landed in Donald Trump's campaign coffers.

Connect those dots. The blueprint is cash, and its intent is clear: neutralize local rules while buying access to a federally hostile regulator.

It's the latest in a string of examples of tech executives cozying up to President Trump's administration, happening as the administration pushes to aggressively back the AI industry and defang state-level regulations that companies like OpenAI have largely opposed. OpenAI did not immediately respond to a request for comment. Brockman's multimillion-dollar donation isn't the only example of him spending big, under his own name, on lobbying efforts that have their sights set on dismantling potential AI industry regulation. The pro-AI super PAC "Leading the Future," of which Brockman is a significant backer, has bought ads targeting New York State Assemblymember Alex Bores -- a cosponsor of New York's RAISE Act, which was watered down at the last minute after coordinated lobbying efforts.

The lobbying blitz worked. New York’s RAISE Act was notably weakened right at the finish line. OpenAI itself?

Silent on its president's political maneuvers. So the dual-track strategy is now operational: fund attacks on lawmakers like Bores, then build a bridge to a sympathetic White House. This isn't a debate.

It's a purchase.

Common Questions Answered

How does GPT-5 differ from previous OpenAI models in terms of system architecture?

GPT-5 is a unified system with three key components: a smart, efficient main model, a deeper reasoning model, and a real-time router that decides which model to use based on conversation complexity and intent. The router is continuously trained on real signals like user model switches and response preferences, with the goal of integrating these capabilities into a single model in the future.

What performance improvements does OpenAI claim for GPT-5 in practical applications?

OpenAI highlights significant advances in reducing hallucinations, improving instruction following, and minimizing sycophancy across three key use cases: writing, coding, and health. The model shows particular strengths in complex front-end generation, debugging larger repositories, and creating responsive websites and apps with an intuitive understanding of design principles.

What are the different versions of GPT-5 that OpenAI is introducing?

OpenAI is launching multiple versions of GPT-5, including gpt-5-main (the primary model), gpt-5-thinking (for deeper reasoning), and mini versions of each model for handling queries when usage limits are reached. Plus subscribers get more usage, while Pro subscribers can access GPT-5 Pro, which offers extended reasoning for more comprehensive and accurate answers.

LIVE02:29Brain Waves Could Guide AI on When to Learn, Neuroscientist Says