Elon Musk has long sounded alarms about artificial intelligence. Now the man who runs OpenAI alongside Sam Altman appears ready to meet him partway. Greg Brockman, OpenAI’s president, has endorsed a proposal from his former colleague for regular gatherings of top AI developers to tackle safety head-on.
The idea isn’t new. Musk first floated industry-wide meetings years ago. Yet the timing carries fresh weight. OpenAI and Musk remain locked in bitter litigation. Brockman’s support, reported in detail by The Information, signals a potential opening amid years of public feuding.
Brockman didn’t mince words. He sees value in structured talks. Companies could share best practices. They might align on risk thresholds before the next leap in model capability. Short meetings. Focused agendas. Real outcomes. That’s the pitch.
History casts a long shadow over this apparent thaw.
Musk co-founded OpenAI in 2015 with Brockman, Altman and others. The nonprofit aimed to steer AI toward humanity’s benefit. Safety sat at its core. Musk pushed hard for caution. He worried about unchecked progress at rivals like Google DeepMind. Then he left the board in 2018.
Testimony from the ongoing trial has laid bare the fractures. Brockman told the court he once feared Musk might physically attack him during a heated exchange over equity, according to NBC News. He questioned Musk’s intuition on AI itself. Early demos left Musk unimpressed. He called the technology “stupid.” Researchers felt deflated. Brockman concluded Musk lacked the depth to judge its trajectory.
Yet here Brockman stands, backing one of Musk’s ideas. Irony abounds. And tension lingers. Musk’s xAI pursues aggressive timelines. OpenAI races forward with GPT successors. Both talk safety. Delivery differs.
Previous efforts offer mixed lessons. The UK government convened global leaders in 2023. Musk attended. Commitments followed on pre-deployment testing. Axios covered the pledges. A second summit in Seoul produced more promises but fewer headlines. Industry groups formed too. The Frontier Model Forum launched in 2023 with OpenAI, Anthropic, Google and Microsoft. Its focus? Coordinated safety research. AI Business detailed the launch.
Those steps fell short for critics. Voluntary. Opaque. No binding rules. Musk has criticized such forums as insufficient. He wants faster action. Independent oversight. Perhaps even pauses on training runs that exceed certain compute thresholds.
Brockman’s endorsement stops short of those demands. It favors dialogue over mandates. Regular meetings could build trust. Share evaluation methods. Discuss deployment guardrails. But skeptics wonder. Will talk translate to slower release cycles? Or just smarter public relations?
Recent developments add urgency. OpenAI formed a new Safety and Security Committee in 2024 as it began training its next major model. Investopedia reported the move. The board, including Bret Taylor and Adam D’Angelo, oversees it. Concerns swirl around capability jumps outpacing controls.
Meanwhile Musk’s legal battle with OpenAI grinds on. He accuses the company of abandoning its founding nonprofit principles. Courtroom drama has revealed old emails and tense all-hands meetings. In one 2018 session, Musk reportedly told staff he saw no path but to race ahead at Tesla. Safety would take a backseat. Brockman read from those records during testimony, per multiple reports including The New York Times.
So why the olive branch now?
Perhaps pragmatism. AI capabilities have surged. Models reason across domains. Multimodal systems handle video, code, complex planning. Risks once theoretical appear closer. Misinformation campaigns. Autonomous weapons. Loss of control scenarios. Regulators watch closely. The European Union passed its AI Act. Washington debates executive orders and legislation.
Industry insiders know fragmentation hurts everyone. A single reckless player could trigger backlash that ensnares all. Coordinated standards might preempt heavier government intervention. Brockman likely recognizes this. Musk too, despite his barbs.
Recent coverage highlights the stakes. A May 2026 report from Tom’s Guide captured Musk calling OpenAI’s current mission a “safety risk” in court. He admitted past mistakes but doubled down on criticism. The trial, though dismissed on some counts, keeps old wounds open.
Yet Brockman’s move suggests pragmatism can cut through animosity. He has steered OpenAI through board upheavals, executive exits and intense scrutiny. His word carries weight inside the company and beyond.
Details on the proposal remain sparse. Frequency. Attendees. Agenda items. All unclear. Would xAI join? Anthropic? Smaller labs? Independent researchers? Inclusion matters. Exclusion breeds suspicion.
Implementation challenges await. Competitive pressures run hot. Talent wars continue. Valuation incentives favor speed. Safety teams at leading firms already collaborate informally. Formalizing that risks leaks or accusations of collusion.
Still. The gesture matters. In an arena defined by rapid iteration, any signal of cooperation stands out. It could seed habits of transparency. Joint red-teaming exercises. Shared benchmarks for dangerous capabilities. Incremental progress.
Watch the response. If Musk amplifies the idea on X, momentum may build. If other CEOs echo support, meetings could materialize within months. Failure to follow through would reinforce cynicism. Another round of lofty talk without teeth.
Either way, Brockman’s endorsement reframes the narrative. Old rivals finding common language on the issue that brought them together originally. Safety. Not as marketing. But as operational necessity.
The AI race shows no signs of slowing. Compute scales. Algorithms improve. Applications multiply. Against that backdrop, industry coordination isn’t optional. It’s table stakes for responsible development. Brockman gets it. Musk has argued it for years. Now the question shifts from whether to meet. To what they will actually achieve when they do.
WebProNews is an iEntry Publication