Leena Rinne: Employees outsource tough talks to AI, creating “socially offloading” spirals

A workplace relationship can break in a way that looks harmless on the surface: an employee gets a message from her boss, assumes it might be AI-written, and asks her own AI tool to decode it. Then, the AI offers a draft response. As Leena Rinne, vice president of leadership, business, and coaching at Skillsoft, describes it, the employee felt like she was stuck watching a strange loop: “‘I literally think [my boss’] AI is talking to my AI. That is the actual conversation happening right now.’” Rinne adds that the employee concluded, “‘I can't crack the code of working with [my boss], because it's just his AI and my AI going back and forth.’”

Rinne has a name for this pattern: “socially offloading.” It’s the workplace version of cognitive offloading, but aimed at the parts of work that require human judgment, empathy, or courage. Similar to shifting menial tasks to technology to reduce mental effort, socially offloading shifts interpersonal skill work to AI. And Rinne’s blunt warning to Fortune is the key: if employees keep asking AI how to respond, “I don’t actually learn how to engage with my boss. I don’t actually learn how to build a relationship with my boss.” In other words, the AI conversation replaces the human one.

To understand why this matters, you have to zoom out from the anecdote to what jobs actually look like day-to-day. Many workplace interactions are “emotion heavy” even when they are business critical: performance reviews, feedback conversations, client updates, and the stressful inbox moments that make people pause before replying. Rinne describes socially offloading in exactly those terms. A boss preparing for a performance review might ask AI how to have the conversation. An employee might ask AI to craft a response to a stressful email from a manager. These are not minor side quests. They are the moments where culture is built, trust is tested, and misalignment gets corrected or hardens into resentment.

Rinne’s argument is not that AI is useless. In fact, she points out the problem is not that AI can’t offer helpful advice. The risk is what the advice costs: the skills that get delayed or never developed. “The risk is then that we don't develop these critical skills that we can use in the moment,” she says, “because we don't know how to navigate emotional intelligence, if AI is navigating emotional intelligence for us.” That is a second-order trap boards and exec teams should recognize. When people stop practicing, their capability stops compounding. The output might look fine in the short run, but the team’s ability to handle ambiguity, conflict, and nuance degrades.

Rinne also frames socially offloading as part of a bigger leadership and organizational shift. She says the price paid for cutting middle management is not the root cause, but rather the leadership vacuum that follows. When organizations flatten their structures and cut out middle managers, mentorship and coaching fall away. In that environment, AI can become a substitute teacher. And substitutes do not replace apprenticeship. Skillsoft, where Rinne is a leadership coach, sells AI tools aimed at teaching real-world conversations. Its product, CAISY, is designed for practice and feedback before important work conversations. The approach is not “here's the answer, here's what you should say.” Rinne says the AI instead teaches people how to develop intrapersonal skills through practice, so they can later execute with judgment rather than a script.

This is where the Meta reference becomes more than a name-drop. The article points to Meta as an example of a strategy that flattens the pyramid while leaning on AI, noting that Meta has cut 25,000 jobs since 2022 and touts an AI team with one boss for every 50 engineers. The source notes that a 25-to-1 employee-to-boss ratio is often seen as an outer limit of span-of-control, but the company is going all-in on AI. The underlying issue is not only who manages, but how people develop. Rinne sees upside in fewer managers being able to lead to quicker decisions and more autonomy. Still, she insists that managers remain needed to turn strategy into results, develop talent, and hold a team together.

The most board-level line in the piece is also the most uncomfortable: Rinne warns that organizations can start treating the span of a leader’s role like it’s a math problem when it is really a capability problem. A team can scale coordination with process. But executing under pressure requires interpersonal competence that cannot be fully delegated. The article connects this to generational dynamics too. Rinne says other generations had decades to learn how to navigate change and organizational dynamics, but now “young people enter the workforce, and they're just thrown into the deep end.” She points out that some blame younger workers for struggling with workplace navigation because they are less social, and she cites Tessa West, a professor of psychology at New York University whose research focuses on communication between employees and bosses. West says those shifts in social behavior can affect performance at work, because early relationships teach skills like negotiation and compromise.

Even romantic relationships do not fill the gap, Rinne argues, because the workplace relationship with a manager is its own training ground. She contrasts her own path, saying she had opportunities to be coached and invested in her development, with what she sees now: Gen Z coming in under the assumption of being ready for the pace of change and the dynamics of navigating it. Rinne says leaders are not equipping younger employees with the tools to communicate effectively and exercise judgment, which lowers competitive advantage when human-centric skills drive success in the AI era. The strategic stake is simple, and it’s not about AI competence. It’s about whether organizations are building the emotional and relational muscle that makes AI output usable in the real world.

For executive teams exploring AI rollout plans, the immediate question becomes: where is your org building the skill, and where is it offloading the practice? Because socially offloading might feel efficient, but it risks turning everyday work relationships into AI-mediated transaction loops, leaving teams less able to negotiate, coach, and align when it matters most.