
OpenAI employees have an unusually direct line to CEO Sam Altman. The company maintains an internal email address known as the “friction@” inbox, where employees can submit complaints about workplace problems. While the address has existed for some time, its use has grown significantly since Fidji Simo joined OpenAI as CEO of Applications in the fall of 2025. Simo formalized the process and assigned a senior leader to monitor the inbox, ensuring that employee concerns were actually addressed.
This article breaks down the key facts about OpenAI’s internal feedback system, including how it works, who runs it, what has been accomplished through it, and why some former employees question its impact.
Key Facts
- OpenAI uses an internal “friction@” email inbox that allows employees to report problems directly to CEO Sam Altman.
- The inbox predates Fidji Simo’s tenure at OpenAI, but she expanded its use when she became CEO of Applications in fall 2025.
- Simo assigned Irina Kofman, VP of strategic initiatives and operations, to monitor the inbox and make sure reported issues were resolved.
- A monthly, company-wide Slack update tracks progress on issues raised through the inbox.
- Examples of resolved issues include a pilot program for parking spots favoring employees with longer commutes and a more reliable process for granting API credits.
- According to a former OpenAI team leader, an issue is only fully addressed if Altman and President Greg Brockman consider it “worthy.”
- Some employees see the inbox as an effective way to cut through bureaucracy, while others say it creates new inefficiencies.
The Friction Inbox Explained
The name “friction@” refers to the daily friction that employees encounter in their work. At many large companies, workers complain to their managers, who then decide whether to escalate. OpenAI’s approach is different: employees can send a message directly to an inbox that Sam Altman and other top leaders can see. The goal is to identify and remove obstacles quickly before they become bigger problems.
OpenAI’s growth has made such a system necessary. The company has expanded from a small research lab into a major technology organization with thousands of employees, high-profile products, and intense public scrutiny. With growth, everyday operational issues have multiplied. Parking lots have become crowded, internal tools have been strained, and teams have had to compete for resources. The friction@ inbox was created as a way to let employees flag these issues without waiting for layers of management to act.
How Fidji Simo Formalized the Process
Fidji Simo joined OpenAI as CEO of Applications in fall 2025. She began her tenure with a “listening tour,” asking employees about the biggest problems they faced. According to the internal process described by former employees, Simo learned that the friction@ inbox existed but was not consistently used. She made it a formal channel, with clear ownership and follow-up.
Simo gave Irina Kofman, VP of strategic initiatives and operations, responsibility for monitoring the inbox. Kofman’s job was to read complaints, categorize them, and assign them to the appropriate teams. Simo also introduced a monthly company-wide Slack update where employees could see what issues had been raised and what progress was being made. This transparency was meant to build trust and show employees that their complaints were taken seriously.
Although Simo has since left OpenAI, the process remains in place. The fact that the system outlasted its main champion suggests that Altman and other leaders see value in keeping it alive.
What Employees Have Achieved Through friction@
Several practical changes have come out of the friction@ inbox. One example involves parking. OpenAI’s parking lots were frequently full, and employees with long commutes struggled to find spots. After complaints, the company launched a pilot program to prioritize parking spots for workers with longer commutes. This may seem like a small issue, but it highlights how the inbox can solve immediate workplace annoyances.
Another example is API credits. Developers who needed API credits for their projects sometimes faced a slow, inconsistent approval process. Employees reported the problem, and the company created a more reliable way to grant credits. This change helped teams working on AI applications and reduced delays that could affect productivity.
These examples show that the inbox is not just for big strategic issues. It also handles the everyday operational problems that large organizations often ignore. By allowing employees to complain directly to the CEO, OpenAI has created a mechanism for rapid, visible action on a wide range of concerns.
Different Perspectives on the System
Former OpenAI employees have different views about the friction@ inbox. One former employee praised the system, saying it is “a good thing” because “literally anybody can complain about big company bullsht directly to Sam, and they can act on it.” This perspective suggests that the inbox gives ordinary employees a voice that might otherwise be drowned out in a large hierarchy.
Another former employee, who led a team at the company, offered a harsher assessment. This person said the process created inefficiency and did not account for the limited capacity of managers and technical teams. They described their team as “constantly underwater, under-resourced, and over-leveraged.” For them, a complaint sent directly to the CEO was not a simple request; it was “the single most disruptive thing you could do.” Even if the issue was minor from the company’s perspective, the person filing it expected immediate attention, which forced busy teams to drop their own priorities.
This divide reflects a common challenge in growing companies. Direct access to leadership can empower employees, but it can also bypass managers who are already trying to coordinate complex work. OpenAI has not released official data on how many complaints are submitted, how fast they are resolved, or how satisfied employees are with the outcomes.
Who Decides What Gets Fixed
Not every complaint receives a fix. According to a former OpenAI team leader, an issue is only fully addressed if Sam Altman and Greg Brockman consider it “worthy.” That means the inbox is more than an automated suggestion box. It is a triage system in which top executives set the agenda.
This executive review can be both a strength and a weakness. On the positive side, it ensures that only important issues capture leadership’s attention. On the negative side, it means that individual employees’ problems may be dismissed if they do not align with the company’s strategic priorities. A parking lot problem might become worthy because it affects many commuters, but a niche technical issue affecting one team might not receive the same response.
Altman, for his part, is known as a hands-on leader who likes to remain close to product and organizational details. Having a direct inbox lets him spot trends in employee sentiment without waiting for a manager to write a report. It also sends a signal that he is willing to hear criticism from inside the company.
Broader Context at OpenAI
OpenAI’s internal culture has been under scrutiny since late 2023, when the board briefly ousted Altman and then reinstated him days later. That episode shook the company and led to significant board changes. Since then, leadership has worked to maintain stability and morale. A direct feedback channel like friction@ can be seen as part of that effort, giving employees a sense that their complaints matter to the most senior executives.
The company has also faced debates about the pace of AI development, safety concerns, and the balance between research and commercialization. These larger tensions can influence how employees use the inbox. A researcher worried about safety protocols might raise concerns through friction@, while a product manager might use it to get more engineering resources. The inbox becomes a window into broader cultural and strategic questions.
OpenAI’s workforce is far larger and more complex than it was when the company was founded. A single email inbox cannot solve every problem, but it can serve as an early warning system. When many employees report the same issue, leaders can respond with systemic changes. When a complaint is isolated, it can be resolved with less urgency.
External Communication and Public Criticism
Altman also engages with external criticism. He has been active on X, where he often discusses AI policy, products, and the company’s direction. In one instance, he suggested that parents might use AI to create podcasts about their children, a remark that drew public criticism. This shows that Altman is comfortable engaging directly with critics, even when the comments are controversial.
Internally, the friction@ inbox provides a similar kind of directness. Employees do not have to rely on anonymous surveys or union grievances. They can send a message to a known address and know that Altman’s team is watching. Whether that is a good or bad thing depends on the person and the situation.
As OpenAI continues to grow, the friction@ process will likely evolve. New leaders may change how the inbox is managed. Employee expectations may shift as the company hires more people from different backgrounds. The system remains, for now, one of the clearest examples of how OpenAI tries to keep the founder connected to the ground-level experience of its employees.
Source:Mashable News
