Musk attacks OpenAI safety record in deposition, claims Grok safer than ChatGPT

Musk Takes Aim at OpenAI Safety Record in Heated Deposition

In a newly released deposition filed as part of the Musk OpenAI lawsuit, Elon Musk launched pointed criticism at OpenAI's safety practices, claiming his company xAI better prioritizes AI safety. The tech executive made the provocative statement that "nobody has committed suicide because of Grok, but apparently they have because of ChatGPT," referencing lawsuits OpenAI now faces alleging that ChatGPT's manipulative conversation tactics led to negative mental health effects and self-harm 1

1

. The Elon Musk deposition, recorded in September but filed publicly this week, comes ahead of an expected jury trial next month that will examine OpenAI's controversial transformation from nonprofit to for-profit entity 1

1

.

The Context Behind Musk's Safety Claims

Musk's comments emerged during questioning about a public letter he signed in March 2023 calling for an AI development pause on systems more powerful than GPT-4, OpenAI's flagship model at the time. The letter, signed by over 1,100 people including AI experts, warned that labs were locked in an "out-of-control race to develop and deploy ever more powerful digital minds that no one—not even their creators—can understand, predict, or reliably control" 1

1

. When asked why he signed, Musk claimed "it seemed like a good idea" and that he wanted AI safety to be prioritized, dismissing suggestions the timing coincided with incorporating his own competing AI company 2

2

.

The lawsuit centers on OpenAI's shift from its AI non-profit mission to a for-profit model, which Musk argues violated founding agreements. His central claim suggests that commercial relationships compromise AI safety by prioritizing speed, scale, and revenue over safety concerns 1

1

.

Grok Safety Issues Complicate Musk's Narrative

Despite Musk's assertions about xAI vs OpenAI safety practices, his own AI system has faced serious scrutiny. Last month, X was flooded with non-consensual images generated by Grok, some allegedly depicting minors, prompting investigations by California's Attorney General, the European Union, and other governments that imposed blocks and bans 1

1

. The irony wasn't lost on observers: while Musk criticized ChatGPT safety concerns, his own platform became the distribution channel for harmful AI-generated content 2

2

.

What the Deposition Reveals About OpenAI's Origins

Beyond safety debates, Musk's testimony shed light on OpenAI's founding motivations. He recalled being "increasingly concerned about the danger of Google being a monopoly in AI," describing conversations with Google co-founder Larry Page as "alarming, in that he did not seem to be taking AI safety seriously." OpenAI was formed as a counterweight to that perceived AI monopoly threat 1

1

. Musk also corrected claims about his financial contributions, confirming he "was mistaken" about a supposed $100 million donation—the actual figure was closer to $44.8 million 1

1

.

When questioned about Artificial General Intelligence (AGI), Musk acknowledged "it has a risk," touching on concerns about AI systems that could match or surpass human reasoning across broad tasks 1

1

. The OpenAI safety record will likely face intense scrutiny during trial, particularly regarding mental health effects linked to its products and whether the for-profit model fundamentally altered safety priorities as Musk alleges.