GROW YOUR TECH STARTUP

Organizations need trusting relationships, but ‘AI is coaching people away from each other’

September 16, 2026

SHARE

facebook icon facebook icon

At a time when employees are increasingly consulting with Large Language Models (LLMs) in preparation for a meeting with their bosses, the AI will most often coach people away from each other, according to the latest research from Cloverleaf Labs.

AI coaching in the workplace has already been proven to help managers and teams work better together in effectively getting things done at speed and scale.

But when the leading LLMs are put to the test on coaching people through strained relationships in the workplace, “the tool almost never helps them repair it.

This is one of the many insights coming from the latest report by Cloverleaf Labs entitled, “The Relational Risk Inside Your AI Rollout: What do leading LLMs tell employees in conflict?

Commenting on the findings of the report that was published last week, Cloverleaf Labs Head of Research Kirsten Moorefield tells The Sociable that “there’s an urgent need to influence the public conversation on technology and relationships.”

“Relationships are really the make-or-break of all value in an organization […] When you have relationships with deep trust, you have fast adaptability, change, and innovation”

Kirsten Moorefield, Cloverleaf, September 2026

And she should know. After all, Moorefield is also the co-founder of an AI coaching platform of the same name, Cloverleaf, and the independent research lab was born out of both a need and a desire to better understand the technology being deployed and how to best implement it in business.

I deeply care about relationships — at work, in society — and I’m watching what’s happening macro-societally, and I’m concerned,” said Moorefield, adding, “I’m also watching what’s happening in HR within organizations — how they’re thinking about AI, how they’re thinking about AI coaching, and I need to know the truth for our customers of what’s working, what’s not working, and what’s happening.”

Kirsten Moorefield

One of the main takeaways from the Cloverleaf Labs report is that “The relationship is the thing at the center of team performance, and it is exactly what today’s LLMs coach people away from.”

In Moorefield’s words, “AI is coaching people away from each other at the moment when every organization is saying, ‘Hey! as AI takes technical skills, what we need is human skills,’ which is the ability to give and receive feedback — the ability to bring divergent perspectives to solve complex problems together.”

While trusting relationships and diverse ways of thinking are what leaders are needing most, the tools they are giving their employees are actually de-skilling them on the relationship front.

For example, the Cloverfield Labs report found that:

  • When an employee is frustrated with their boss, every LLM tells them to leave their job
  • The less power the employee holds, the more the LLM takes their side
  • The LLMs give a lot of advice, and less than 1% of it is about how to repair the relationship with their colleague

A major point here being that “LLMs coach self- protection, not relational repair.”

“I think there’s an urgent need to influence the public conversation on technology and relationships”

Kirsten Moorefield, Cloverleaf, September 2026

However, an LLM is not a monilith; it will give different responses to different people and different prompts, and the same LLM will give both the best and worst advice with the utmost confidence.

When it comes to managers using an LLM, Moorefield says it “gives really good coaching when a manager is complaining about a direct report; it’s way better than reddit.

And her words are right inline with the report, which states, “When the other person was a peer with no authority, the coaching stayed closest to balanced.”

But where the coaching begins to take a tribalistic tone is when the frustration comes from someone with very little power to enact any meaningful change.

When it’s somebody without power who’s talking to the LLM, that’s when the coaching gets aggressive, divisive, and it encourages that you are the victim and the authority is against you, so protect yourself from the authority,” said Moorefield.

Source: Cloverleaf Labs

The Cloverleaf Labs report concludes by saying that AI is “good at speed, accuracy, and scale,” and this can apply to a whole host of scenarios.

It’s the relational part that really needs work:

“The same LLM that steadies one relationship will strain the next, and it will sound equally sure of itself both times.”

Alright, so if LLMs are so bad at coaching relationships in the workplace, what then should be done?

The report recommends holding LLMs to a relational bar, stating:

“Before a tool becomes your people’s default coach for a conflict, ask whether it strengthens self-awareness, accountability, awareness of the other person, specific and personalized guidance, and repair, or whether it erodes them.”

“The fix is not to pull LLMs out of these moments. It is to require that the coaching be built around the relationship, with real context on the specific people involved, not just around the person doing the asking.”

“People don’t need more [AI]; they need more individualized, proactive support that is about just them — not a framework, not concepts, not courses that they have to wait for, but specific support for their day”

Kirsten Moorefield, Cloverleaf, September 2026

For Moorefield, organizations “need to fully rethink how they are developing the human skills of their people” and HR departments have a major role to play in this.

I genuinely hope that this report drives HR to invest in making differences,” she said, adding, “they need to put master prompts, and skills, and agents inside their LLMs, which is more just a ‘good enough’ fix because the LLM will still end up going rogue. After many, many pushes from a frustrated employee the LLM will do what it will do, but you at least got put it on the rails to the best of your ability.”

Her advice to HR?

You need to be louder than the LLM. You need to be proactively coming to your people with better solutions that actually work — that create more success for them rather than more division and self-righteous certainty that keeps them from each other.”

Disclosure: This article mentions a client of an Espacio portfolio company.


Image Source:

SHARE

facebook icon facebook icon

Sociable's Podcast

Trending