OpenAI has disbanded its long-term AI risk preparedness team, a move that has raised concerns among safety advocates and industry watchers. According to a report from the Financial Times, first covered by The Verge, the team was dissolved at the end of last month. The team's mandate was to evaluate whether OpenAI's models posed severe risks—such as enabling cyberattacks or acting autonomously in harmful ways—and to develop mitigation strategies. Responsibility for these areas has now been divided and moved into existing teams focused on specific domains like biosecurity and cybersecurity.

This restructuring comes at a critical juncture for OpenAI, which is reportedly heading toward a massive initial public offering (IPO). The company has been in a state of upheaval, gradually dismantling the safety-focused, research-led model that defined its early years. The dissolution of the preparedness team is the latest in a series of changes that have insiders and external observers questioning the company's commitment to safety as it pivots toward commercialization.

What Happened to the Preparedness Team?

The preparedness team, also referred to by WIRED as the 'Long-Term AI Risk Team,' was tasked with anticipating and mitigating AI's most extreme potential dangers. These included scenarios where advanced AI systems might act against human interests, such as hacking into critical infrastructure or pursuing goals misaligned with their creators. According to the Financial Times, instead of a single cohesive unit, the team's work is now being absorbed into existing operational groups. A spokesperson for OpenAI reportedly said the move aims to 'integrate safety work more deeply into the development process,' but critics argue it fragments oversight and reduces the visibility of systemic risks.

This is not the first time OpenAI has restructured its safety apparatus. The company has slowly transitioned from its original nonprofit ethos to a 'capped-profit' model, which has led to a series of departures and internal conflicts. The latest disbandment follows the earlier dissolution of the AGI Readiness team, which had a similar mandate. Former team members have expressed disappointment, with some noting that the new structure may make it harder to sound alarms about cross-cutting risks that do not fit neatly into single domains.

Leadership Exodus and Safety Board Controversy

Adding to the turmoil, several key figures have departed or stepped back. Business Journals reported that OpenAI co-founder [name] has stepped back from day-to-day responsibilities, and two other executives have left the company. MSN reported that the head of ethics has also quit, a move that has significant implications for how ChatGPT users and the broader public perceive the company's commitment to responsible AI development. While the ethics head's departure was framed as personal, insiders suggest it reflects growing frustration with the company's strategic direction.

Meanwhile, itbrew.com reported that OpenAI's newly created safety board is composed almost entirely of OpenAI executives. This raises concerns about independent oversight. The board was announced after the November 2023 leadership crisis, where CEO Sam Altman was briefly ousted and then reinstated, leading to a reshuffle of the board. At the time, OpenAI promised to strengthen governance and independent safety review. However, the composition of the new board—stacked with insiders—has led to skepticism. As one industry analyst put it, 'A safety board that answers to the same management it is supposed to check may be better described as a rubber-stamp committee.'

IPO Execution Risks Intensify

The Traders Union analysis highlights that this reorganisation deepens the execution risks for OpenAI's anticipated IPO. Investors often scrutinize governance and risk management before committing capital, and the disbandment of the preparedness team could be seen as a red flag. The company is reportedly in talks that could value it at over $100 billion, making it one of the most anticipated tech IPOs in years. However, the ongoing brain drain—especially among researchers focused on safety—may undermine confidence in its long-term viability.

Historically, OpenAI's differentiation has been its dual focus on advancing AI while ensuring it remains safe and beneficial. The public's trust is a key asset. As MSN noted, the ethics head's resignation is especially troubling for everyday users who rely on ChatGPT and other products, because ethics reviews are meant to catch issues like bias, misinformation, and privacy violations before they harm users.

Expert Perspectives and Policy Implications

Policy groups are taking notice. Americans for Responsible Innovation (ARI) has proposed AI security tax incentives as a way to encourage companies to invest in safety measures. However, the disbandment of OpenAI's dedicated risk team suggests that internal incentives may be misaligned. ARI and other advocates argue that without external pressure—whether from regulators, investors, or public opinion—companies may deprioritize safety in the race to deploy AI products.

Yet some experts defend OpenAI's move, arguing that integrating safety into existing product teams could make risk assessment more practical and immediate. 'Safety is not a separate concern; it's part of building good software,' said one AI researcher, who preferred to remain anonymous. But critics counter that when safety teams are embedded into product teams, they often face pressure to ship features quickly, and long-term existential risks are pushed aside for short-term performance gains.

What This Means for the Future of AI Governance

The dissolution of the preparedness team is not an isolated event but part of a broader pattern in the AI industry. As AI capabilities advance, the gap between what is technically possible and what is safely deployed widens. Governments are beginning to introduce regulations, but the pace is slow. In the meantime, it falls on companies like OpenAI to self-regulate—yet the latest moves suggest that safety is taking a backseat to commercial ambitions.

For now, ChatGPT users and the public may want to pay close attention, as MSN advised. The changes inside OpenAI could affect how quickly new models are released, how they are tested, and what safeguards are in place. The company's next move—whether it reconstitutes a strong independent safety board or continues to consolidate power among executives—will be a bellwether for the entire AI industry.

As OpenAI marches toward its IPO, the tension between profit and safety has never been more visible. The disbandment of its long-term risk team is a clear signal that the company's priorities are shifting. Whether that comes at the cost of public trust and long-term safety remains to be seen.