Reports of internal reporting at OpenAI on an innovation to develop powerful artificial intelligence models have furthered speculations around the initial dismissal of CEO Sam Altman, who was reinstated four days later.
The model, Q*, made a breakthrough in its ability to solve math problems. Experts see it as a step toward creating artificial general intelligence (AGI), autonomous systems that may surpass humans in most economically valuable tasks. AGI can generalize, learn, and comprehend. AI experts told Business Insider that it represents a symbolic leap from existing models, which have struggled to generalize outside the data they are trained on.
According to a Reuters report citing two sources, multiple OpenAI researchers and employees wrote a warning to the company’s former Board of Directors expressing safety concerns about Q*, flagging the potential dangers.
Commercial behavior with risky technology poses a significant risk to investors and the public.
Q* was only one of the several issues the group of employees who wrote the letter to the Board had with how OpenAI was being run, specifically over commercializing advances before understanding the consequences.
The Board had cited Altman’s “miscommunication” and “lack of transparency” in their initial decision to remove him. In a blog post on the day of his dismissal, November 17, the company stated: “He was not consistently candid in his communications with the Board, hindering its ability to exercise its responsibilities.”
Though deliberately vague in language, the Board of Directors alleged false statements.
Whistleblower experts recognize that this behavior of miscommunication and lack of
transparency may point to instances of false claims, misleading investors, or the result of an
The letter only added to a long list of grievances the board had against Altman, according to Reuters’ sources.
Employees can anonymously raise concerns about false statements and internal control. Internal reporting becomes information valuable for enforcement action when brought to the correct agency even in cases of restrictive NDAs and company self-reporting under Dodd-Frank
Whistleblower attorney Stephen M. Kohn, founding partner at Kohn, Kohn & Colapinto and Chair of the National Whistleblower Center stated that “If any insiders have evidence to reasonably believe that the content of the ‘misleading communication’ cited by the Board violated the law, they should file an anonymous report as soon as possible.”
If, as alleged, Altman was in fact not transparent with the Board, it is plausible that OpenAI might have made materially false and misleading statements to the public, which has led to SEC enforcement actions in numerous cases.
An important detail, however, is that due to the nonprofit structure of the Board, they do not represent investors in their actions. Investors’ threat to sue the Board of Directors preceded the reinstatement of Altman and subsequent dismissal of Board Members Helen Toner and Tasha McCauley reinforces that. Hence, Altman’s misleading communication may not be considered “misleading investors” as a technicality.
Kohn urges those with information about potential fraud in the under-regulated industry to inform the appropriate agencies. Whistleblowers can receive monetary rewards for anonymously providing information that protects the public from harm caused by a potential fraud scheme. Reports can be made to the Securities & Exchange Commission and Commodity Futures Trading Commission under the Dodd-Frank Act.
Under the SEC Whistleblower Program, for example, whistleblower tips that lead to a successful enforcement action may be entitled to an award of 10-30% of the sanctions collected by the government. They could still qualify for an award even if the company has self-reported the incident.
Kohn stated, “It is important for employees of OpenAI to know that they have the option and incentive to report wrongdoing, even if their company tells them otherwise.”