Greg Brockman, President of OpenAI, recently shed new light on internal turmoil at the company, detailing a heated confrontation with Elon Musk and raising fundamental questions about artificial intelligence governance.
What happened
Brockman testified in court, describing a tense meeting with Musk that preceded his departure from OpenAI's board in 2018. Musk, an OpenAI co-founder, had expressed strong concerns about the company's direction and its transition to a "capped-profit" model, believing it deviated from the original non-profit mission. The testimony, reported by Wired AI, highlighted how strategic and personal differences led to a highly strained atmosphere, culminating in Musk's demand to take full control of the company or see it fail. Brockman also mentioned subsequent attempts to remove other board members, including Helen Toner and Tasha McCauley, amid growing internal polarization.
Why it matters
The internal affairs of a company like OpenAI are not mere corporate gossip; they touch the core of AI governance and public trust. OpenAI is a dominant player in the development of generative AI models, and its decisions influence the entire technological and social ecosystem. A lack of clarity and stability in leadership can undermine the confidence of users and regulators, especially in a sector where transparency and accountability are crucial. These conflicts highlight the difficulty of balancing rapid innovation, commercial objectives, and the ethical mission of developing AI that benefits humanity. When leaders of these organizations are embroiled in power disputes, the question arises of who holds true control and with what priorities. This directly impacts the perception of ethical AI development, making it harder for the public to trust the safeguards implemented.
The HDAI perspective
From the Human Driven AI perspective, the revelations about the Brockman-Musk conflict underscore a critical point: the governance of major AI organizations is as important as the technology itself. It is not enough to develop powerful models; it is essential that decision-making structures are robust, transparent, and aligned with clear ethical principles. OpenAI's transition from a non-profit to a for-profit entity, while retaining an original mission, created inherent tensions that continue to manifest. These episodes reinforce the need for open and structured debate on the responsibility of tech companies and external oversight. Topics such as the necessity of ethical AI and transparent governance will be central to the HDAI Summit 2026 in Pompeii, where experts and stakeholders will discuss how to build an AI future that prioritizes humanity. In parallel, research continues to explore technical solutions to improve AI reliability, as demonstrated by the ERA (Evidence-based Reliability Alignment) framework, which aims to enhance the honesty of RAG (Retrieval-Augmented Generation) systems by managing knowledge conflicts between models and retrieved information ArXiv cs.AI. This technical approach to AI "honesty" is a necessary complement to robust human governance.
What to watch
The implications of these legal and internal disputes will continue to shape the AI landscape. It will be crucial to observe how OpenAI manages its image and the trust of partners and users. Discussions about the future of regulation, such as the EU AI Act, could gain further momentum from cases like this, which highlight the risks of concentrated and insufficiently controlled power. The ability to balance innovation with responsibility will remain the central challenge for the entire industry.