Many of us, not least in HR, are grappling with how to use artificial intelligence (AI) across the workplace. The mainstream belief, or hope, is that AI will make work easier and more efficient, and so increase productivity. But it’s also important to consider its impact (positive or negative) on workplace relationships.
With AI, are we missing the point?
Blind faith in technology, pressure from social media and worries that the firm may be ‘left behind’ all direct attention away from a complex and yet crucial question: How will AI adoption affect workplace relationships?
As it stands, many organisations neglect relationships. Managers lacking interpersonal skills rely on a rule book. Inadequate or outdated systems reinforce silos. Colleagues are too busy or stressed to talk with each other. Pursuit of near-term outcomes encourages ‘transactional’ exchanges.
While mechanistic thinking about performance is the norm, its day-to-day practice hurts experiences, productivity and results. Modern work demands that people collaborate on complex problems: no brandishing of managers’ whips recovers potential lost to bureaucratic methods.
Whether corporate motives behind the adoption of AI are good or doubtful, you have the freedom to protect your workplace relationships..
AI and workplace relationships
If technology is to help rather than harm, it must amplify and not muffle the human relationships that make cooperation possible. To evaluate AI against this yardstick, let us examine several ways in which platforms are, or may be, used across the workplace.
1. Freedom from drudgery
AI, apologists say, will pick up the drudgery and liberate you for what matters most, tasks only humans can do. Relationships demand time and energy so less effort spent on tedious activities is clearly a benefit.
Still, evidence suggests that firms see AI as a way to cut labour costs, not as a means to augment roles. For example, when the online payments company Klarna introduced AI, the shift spelled much gloom and little boon for thousands of employees.
Automation, of course, heralds the possibility of more time for enriching work. But the Silicon Valley hype is misleading: you cannot connect after being fired from a job. Likewise, fear of being ‘replaced with AI’ urges a scramble to appear useful, where relationship-building typically loses out to short-term priorities.
2. Presentation
A town hall keynote in English may be translated, via deep-fake technology, into a convincing rendition in (say) Japanese, with matching tone and body language. This would allow you to explain the company’s 2025 vision for a global audience.
Some employees may find such a deep fake inauthentic and off-putting; they may prefer, and expect, to see you on stage, albeit at the cost of an imperfect translation. Others may welcome the clarity of the spoken message and, through a realistic if not real video, connect with you as the original speaker.
3. Email
Recent ads for Apple Intelligence showed how AI may be used to write email responses with little or no intervention. LinkedIn users found this a way to satisfy social norms for being ‘polite’ when one could not ‘be bothered’ to write a well-rounded message.
Many of us suffer from email overload, and AI-generated text may feel attractive. But beyond basic transactional emails – ‘Yes, I am free to meet at 2:30 p.m.’ – messages are a chance to connect with another person. AI-embellished emails likely create a false impression, perhaps deliberately, and this harms trust.
Some colleagues may be happy to receive an AI-generated message. Others may see you as disrespectful and prefer no response, a hurried reply or even a vexed one, if this mirrors the reality of the situation or your relationship.
4. Feedback
Generative AI makes it quick and easy for you to give someone feedback, be it a routine appraisal, comment on a piece of work or simply a piece of advice.
But to invite or receive feedback is an act of vulnerability. It is also a mark of respect, for your experience, opinion or insight. A colleague who asks for feedback will not care about AI’s ‘views’ – it has none – or the embellishments a large language model (LLM) can make to your own ideas.
If authenticity and trust are to be maintained, feedback is one area where you are advised to avoid AI.
5. Learning
Much workplace conflict can be avoided. As I set out in two articles on ‘difficult’ colleagues – those who typically either think differently or act badly – we can all learn to build more productive relationships.
Here, a bot can role-play the words and behaviours of an intractable and yet fictitious team member, giving you feedback and guidance on how to navigate relationship obstacles. AI-assisted learning thus may be a safe means to acquire management and/or interpersonal skills.
Note that in this constructive use of AI, the technology does not mediate a live relationship with another person. Friction is more likely to arise where AI sits between you and other people.
Protecting workplace relationships: Three imperatives
AI is here to stay, and organisations will find, or make, uses for the technology. Whether corporate motives behind the adoption of AI are good or doubtful, you have the freedom to protect your workplace relationships.
Consider these three imperatives:
First, a relationship always involves two people, so remain mindful of how the other party might respond to AI-assisted interaction. Some people enthuse over the technology; others have a visceral reaction against it.
Second, think about trust as a guiding principle. In any given interaction, will trust be built or breached if you make use of AI? An AI translation may enable a faster and so welcome response, whereas LLM-inspired compliments may cause offence.
Finally, examine your own motives for using AI, and consider how the other person might feel when he or she realises what these say about you or your relationship. With human connection, efficiency is rarely the point; and you cannot outsource the building of trust.
