Elon Musk’s only AI expert witness at the OpenAI trial fears an AGI arms race

The News

The legal battle between Elon Musk and OpenAI took a dramatic turn this week as Stuart Russell, Musk’s sole expert witness, raised concerns about a potential "AGI arms race" [1]. Russell, a renowned AI researcher and author of Artificial Intelligence: A Modern Approach, testified during the trial, which centers on Musk’s claim that OpenAI deviated from its original mission of developing AI for humanity’s benefit [1]. Musk alleges that OpenAI, now a for-profit entity, prioritizes rapid advancement and commercialization over safety and ethical considerations [3]. The trial, expected to last several weeks, has already revealed contentious details, including text messages where Musk warned OpenAI’s leadership they would become "the most hated men in America" if they didn’t settle the lawsuit [3]. This testimony and the broader trial proceedings underscore a growing tension over the trajectory of advanced AI development and the risks of unchecked progress [1].

The Context

The lawsuit stems from Musk’s initial involvement in OpenAI’s founding as a non-profit research organization in 2015 [1]. Musk contributed approximately $38 million to the organization [2, 3], envisioning a collaborative effort to ensure AI’s safe and beneficial development [1]. However, Musk alleges that OpenAI’s shift to a capped-profit model and its pursuit of powerful AI models, including Sora, represent a betrayal of this mission [1, 4]. The for-profit structure, while attracting investment and talent, introduced a commercial imperative Musk believes has overshadowed safety concerns [1]. This shift reportedly led to Musk’s departure from OpenAI’s board in 2018 [1].

Russell’s concerns about an AGI arms race stem from the competitive landscape, particularly the rivalry between OpenAI, Microsoft (a major investor), Google (via DeepMind), and Musk’s xAI [1, 4]. The pursuit of Artificial General Intelligence (AGI)—defined as AI capable of understanding, learning, and applying knowledge across tasks at a human level or beyond—is driving this competition [1]. Models like Sora, which generate realistic video content from text prompts, exemplify this rapid advancement [4]. xAI, founded by Musk in 2023, is explicitly focused on developing AI to understand the universe’s nature [4]. Musk has stated that xAI’s models are designed to "distill" OpenAI’s models, indicating a competitive strategy of reverse engineering and accelerating progress [4]. Russell argues this pressure incentivizes labs to prioritize speed over safety, risking a dangerous escalation in AI capabilities without safeguards [1]. OpenAI’s current valuation is estimated at $800 billion, with some projections reaching $1.75 trillion [2, 4], further fueling the competition [4].

The trial also revealed that Musk’s $38 million investment represents a small fraction of OpenAI’s valuation [2]. This disparity highlights the financial incentives driving the company’s trajectory and potential conflicts between Musk’s concerns and OpenAI’s commercial goals [2]. Musk’s testimony included stumbles, such as admitting he didn’t know the definition of a key technical term [2], raising questions about the strength of his legal arguments and strategy to oust CEO Sam Altman and revert OpenAI to a non-profit status [2].

Why It Matters

The potential for an AGI arms race, as highlighted by Russell’s testimony, has significant implications for developers, enterprises, and the AI ecosystem. For engineers and researchers, the pressure to rapidly innovate and deploy powerful models creates technical friction [1]. The focus on achieving AGI often necessitates cutting corners on safety testing and ethical considerations, risking unintended consequences and model failures [1]. The rapid pace of development also demands continuous upskilling, straining the existing AI talent pool [1]. Enterprises adopting advanced AI models face risks like data security breaches, bias, and accountability issues, alongside high implementation and maintenance costs [1].

The trial has created a clear divide within the AI community. OpenAI’s leadership, particularly Altman and Brockman, faces scrutiny over potential removal from their roles if Musk’s lawsuit succeeds [2]. This uncertainty has caused anxiety among employees and investors [2]. Conversely, Musk’s actions are criticized as opportunistic, driven by a desire to control AI’s direction [1]. The emergence of xAI as a direct competitor has intensified the race, accelerating AGI development but also escalating risks [4]. The text messages revealing Musk’s prediction that Altman and Brockman would become "the most hated men in America" underscore the high stakes and reputational risks [3].

The legal proceedings are also shaping public perception of AI safety and governance. The trial has amplified concerns about advanced AI risks and the need for oversight [1]. The debate over OpenAI’s for-profit structure highlights the challenge of balancing innovation with ethical responsibility [1]. Open-source models like gpt-oss-20b (6,981,799 downloads) and gpt-oss-120b (4,237,999 downloads) offer alternatives, democratizing access to AI but complicating safety and alignment efforts [1].

The Bigger Picture

The Musk vs. OpenAI trial reflects a broader trend in AI: the relentless pursuit of AGI and the commercialization of research [1, 4]. This trend mirrors actions by other players, such as Google’s DeepMind, which is also investing heavily in AGI [1]. The competition between labs is accelerating AI capabilities but also creating instability and risk [1]. Sora’s development, for example, showcases rapid progress in generative AI but raises concerns about misuse, such as deepfakes and misinformation [4].

The trial’s focus on OpenAI’s for-profit model highlights a debate over AI governance structures [1]. While for-profit models attract investment and talent, they may conflict with safety goals [1]. xAI’s emphasis on fundamental research over commercial applications represents a different approach [4]. Tools like the OpenAI Downtime Monitor (freemium pricing, tracking API uptime) reflect growing awareness of transparency and accountability in AI systems [1]. Research on adaptive attacks on image watermarks underscores the need for robust security measures to protect AI from malicious actors [1].

Daily Neural Digest Analysis

The mainstream media has largely framed the Musk vs. OpenAI trial as a personal feud between tech titans [1]. However, the trial exposes a deeper tension: the conflict between rapid innovation and safety/ethical alignment [1]. Russell’s warning about an AGI arms race is a direct consequence of the competitive dynamics driving AI development [1]. The trial’s revelations about Musk’s initial involvement and disillusionment highlight challenges in maintaining a shared vision for AI’s future [1]. xAI’s active "distilling" of OpenAI’s models [4] demonstrates the intensity of this competition and the risk of a safety standard race to the bottom.

The hidden risk lies not just in AGI development but in its potential to occur without adequate societal consideration [1]. The current regulatory landscape for AI is inadequate, and the trial underscores the urgent need for oversight and accountability [1]. The question remains: can the AI community and the world harness AI’s transformative potential while mitigating existential risks? The trial’s outcome may offer insights, but a fundamental shift in AI development and governance is ultimately required [1].

---

References

[1] Editorial_board — Original article — https://techcrunch.com/2026/05/04/elon-musks-only-expert-witness-at-the-openai-trial-fears-an-agi-arms-race/

[2] Ars Technica — Elon Musk's 7 biggest stumbles on the stand at OpenAI trial — https://arstechnica.com/tech-policy/2026/04/elon-musks-7-biggest-stumbles-on-the-stand-at-openai-trial/

[3] TechCrunch — Elon Musk sent ominous texts to Greg Brockman, Sam Altman after asking for a settlement, OpenAI claims — https://techcrunch.com/2026/05/04/elon-musk-sent-ominous-texts-to-greg-brockman-sam-altman-after-asking-for-a-settlement-openai-claims/

[4] MIT Tech Review — Musk v. Altman week 1: Elon Musk says he was duped, warns AI could kill us all, and admits that xAI distills OpenAI’s models — https://www.technologyreview.com/2026/05/01/1136800/musk-v-altman-week-1-musk-says-he-was-duped-warns-ai-could-kill-us-all-and-admits-that-xai-distills-openais-models/