Why Social Robots Should 'Resist' You Instead Of Just Agreeing With You

We have all seen the future of companionship in a lot of sci-fi movies, but the reality is often a little dull. Most social robots today are designed to be obedient and predictable. They follow commands and prioritize user comfort. But a new study from the University of Bristol suggests that maybe we need a change. The research presented at the CHI '25 conference in Yokohama challenges the idea that robots should simply be passive companions for therapy. Instead, the study suggests that therapeutic robots should act more like active partners.

The inspiration comes from an unlikely source. The study looks at Equine-assisted interventions or EAIs. These are used to help patients with PTSD, trauma, and autism. These individuals often struggle to express and regulate emotions through words alone. The researchers looked closely at how horses work. In EAIs, individuals communicate with horses through body language and emotional energy. If someone is tense or unregulated, the horse resists their cues. When the person becomes calm and clear, the horse responds positively. This creates a "living mirror" effect.

Lead author Ellen Weir from the Faculty of Science and Engineering explains that this is what robots should mimic. Weir noted that EAIs can be expensive and inaccessible due to the need for highly trained horses and facilitators. This is where robots come in. Weir found that therapeutic robots should not be passive companions but active co-workers. Just like horses, therapeutic robots should resist engagement when users are stressed or unsettled.