Inside the Crisis of Trust at OpenAI: Can Sam Altman Lead AI’s Future Responsibly?

In a paradox that has left industry observers reeling, OpenAI—arguably the world’s most influential artificial intelligence company—simultaneously published sweeping policy recommendations designed to ensure AI benefits humanity while a bombshell investigation in *The New Yorker* raised serious questions about whether its CEO, Sam Altman, can be trusted to uphold those ideals. The juxtaposition underscores a growing crisis of confidence at the heart of AI’s future, one that pits OpenAI’s public-facing commitment to safety and equity against a pattern of alleged manipulation and self-interest that has alienated insiders and eroded public trust.

• OpenAI’s new AI policy recommendations aim to prioritize human welfare during the transition to superintelligence, but internal sources and *The New Yorker* investigation cast doubt on CEO Sam Altman’s trustworthiness.
• Insiders describe Altman as a manipulative figure who prioritizes personal power over safety, with internal communications and former leaders alleging repeated deceptions.
• Public skepticism toward AI is rising, driven by concerns over job displacement, child safety, and the environmental impact of data centers, complicating OpenAI’s efforts to shape AI governance.

Why OpenAI’s Policy Vision Collides With Leadership Doubts

On June 25, 2024, OpenAI released a 38-page report outlining an “industrial policy for the intelligence age,” a bold manifesto that envisions a future where AI-driven economic growth is shared equitably through mechanisms like a public wealth fund and shorter workweeks. The document calls for global cooperation to manage risks, including the potential for superintelligent systems to evade human control—a scenario the company acknowledges could “harm people” if unchecked. OpenAI’s chief global affairs officer, Chris Lehane, framed the recommendations as a direct response to growing public anxiety, telling *The Wall Street Journal* that the company is “urgently concerned” about negative perceptions of AI.

A Vision of Shared Prosperity—and Who Would Benefit

Among OpenAI’s most ambitious proposals is a public wealth fund designed to distribute AI-driven profits directly to citizens, ensuring that the benefits of superintelligence are not concentrated in the hands of a few investors. The company also advocates for a 32-hour, four-day workweek with no loss in pay, arguing that reduced labor hours could be reinvested into paid time off or permanent reductions in work hours without sacrificing productivity. To cushion the blow of automation, OpenAI suggests policies like taxing automated labor to fund core social programs—including Social Security, Medicaid, and housing assistance—as companies increasingly replace human workers with AI systems.

The company’s recommendations extend to worker retraining, with a focus on transitioning displaced workers into care-centric professions such as healthcare, elder care, and childcare—roles historically undervalued and underpaid. OpenAI argues that recognizing caregiving as “economically valuable work” could help shift societal attitudes and attract more workers to these critical fields. Additionally, the report proposes incentives for employers and unions to pilot shorter workweeks and for public-private partnerships to accelerate AI innovation while maintaining democratic oversight.

The New Yorker Investigation: A Portrait of Deception and Power

In stark contrast to OpenAI’s idealistic policy blueprint, *The New Yorker*’s investigation—based on interviews with more than 100 people familiar with Altman’s leadership, internal memos, and 12 one-on-one sessions with the CEO—portrays a leader whose actions have repeatedly undermined trust. Former OpenAI board member Helen Toner characterized Altman as possessing “two traits that are almost never seen in the same person: a strong desire to please people, to be liked in any given interaction, and an almost sociopathic lack of concern for the consequences of deceiving them.” The investigation found no smoking gun, but it documented an “accumulation of alleged deceptions and manipulations” that former chief scientist Ilya Sutskever and former research head Dario Amodei concluded made Altman unfit to foster a safe environment for advanced AI development.

“The problem with OpenAI,” Amodei wrote in an internal message, “is Sam himself.”

Contradictions and Shifting Narratives

Altman has responded to the investigation by disputing specific claims, attributing others to his conflict-avoidant nature, and arguing that his evolving positions reflect the rapidly changing AI landscape. Yet these contradictions have become harder to dismiss as scrutiny of OpenAI intensifies. Once known for warning of AI doomsday scenarios, Altman has recently adopted a tone of “ebullient optimism,” even as lawsuits label OpenAI’s technology as unsafe and governments increasingly rely on its models. His shifting stances—from advocating for strict safety controls to embracing rapid deployment—have fueled skepticism about whether OpenAI’s policy recommendations are genuine efforts to address public concerns or strategic maneuvers to deflect criticism.

Public Trust in AI Eroding Amid Safety and Ethical Concerns

The timing of OpenAI’s policy rollout is no coincidence. A Harvard/MIT poll cited by *Axios* in May 2024 found that Americans’ biggest concern about AI is its potential to harm their quality of life, driven by fears of job displacement, child safety risks, and the environmental toll of energy-guzzling data centers. These concerns are gaining political traction, with some local governments considering moratoriums on data center construction that could slow AI advancement. A loss of Republican control of Congress in the 2024 elections could pave the way for stricter AI safety regulations—a prospect that Altman has privately lobbied against, according to *The New Yorker*.

The Data Center Dilemma: Powering AI vs. Public Backlash

OpenAI’s policy recommendations acknowledge the environmental and social costs of AI infrastructure, including the massive energy consumption of data centers. While the company proposes public-private partnerships to address these challenges, critics argue that its push for rapid AI expansion contradicts its stated commitment to sustainability. Data centers, which can consume as much electricity as small cities, have become a flashpoint in communities nationwide, with residents and local officials raising alarms about water usage, grid strain, and environmental degradation. OpenAI’s call for “common-sense regulations” may struggle to gain traction if the public perceives the company as prioritizing growth over accountability.

The Safety Paradox: Can OpenAI Police Itself?

Central to OpenAI’s vision is the idea that advanced AI models should be subjected to rigorous audits—but only for the most capable systems, to avoid stifling competition. The company argues that a global network should be established to communicate emerging risks, with public input playing a vital role in shaping AI governance. However, this self-regulatory approach has drawn skepticism from experts who question whether a company led by Altman—whose leadership has been criticized by insiders—can credibly oversee its own safety protocols. The tension between OpenAI’s idealistic policy goals and its internal culture of opacity raises a critical question: Can a company that has repeatedly prioritized its own interests over transparency and safety be trusted to guide the future of AI?

Altman’s Leadership Style: Charming Pitchman or Unaccountable Power Broker?

Described by *The New Yorker* as “the greatest pitchman of his generation,” Altman has a history of persuading a tech-skeptical public that his priorities align with theirs—even when they are mutually exclusive. Yet insiders and external critics alike argue that his leadership is defined by a pattern of evasion, contradiction, and personal ambition. One OpenAI researcher told *The New Yorker* that Altman’s promises often function as “a stopgap to overcome criticism until he reaches the next benchmark,” whether that’s launching a new product, securing regulatory approval, or outmaneuvering competitors. This approach, the researcher suggested, allows Altman to “set up structures that, on paper, constrain him in the future”—only to dismantle them when they become inconvenient.

The Elon Musk Precedent: A Warning from a Former Ally

Altman’s leadership style has drawn comparisons to his tumultuous tenure at OpenAI, which ended in 2018 when Musk publicly criticized him and exited to launch his own AI venture, xAI. Musk’s departure followed reports of internal conflicts and Altman’s alleged resistance to safety oversight, themes that resurface in *The New Yorker*’s investigation. While Musk’s criticisms may carry personal bias, his warnings about Altman’s approach to AI governance underscore a broader unease about OpenAI’s direction. Some experts estimate that superintelligence—if it arrives—could emerge within the next two years, a timeline that mirrors the brief period Musk spent at OpenAI before his falling out with Altman.

OpenAI’s Response: Idealism Meets Reality

OpenAI has framed its policy recommendations as “initial ideas for an industrial policy agenda to keep people first during the transition to superintelligence.” The company acknowledges that it doesn’t have all the answers but insists its proposals are a starting point for collaboration with governments, workers, and the public. In a statement, OpenAI emphasized its commitment to “common-sense regulations” and a public-private partnership model that balances innovation with safety. However, the company’s ability to deliver on these promises hinges on whether Altman’s leadership can overcome the doubts sown by *The New Yorker*’s investigation and restore credibility with insiders, regulators, and the public.

The Broader Implications: AI Governance in an Era of Distrust

The crisis at OpenAI reflects a larger reckoning within the AI industry, where trust is increasingly scarce and the stakes could not be higher. As AI systems grow more capable, the risks of misuse, unintended consequences, and unequal access become more acute. OpenAI’s policy recommendations—while ambitious—offer little concrete guidance on how to enforce accountability or prevent the concentration of power in the hands of a few dominant firms. Without robust external oversight, there is a real danger that AI’s benefits will be unevenly distributed, its harms inadequately addressed, and its development driven by the whims of unaccountable leaders like Altman. The question now is whether OpenAI’s vision can survive the man at its helm—or if the company’s ideals will be subsumed by the very forces it claims to resist.