These findings suggest that robot control does not uniformly affect moral behavior, but can selectively shape social and prosocial decisions.

Participants remotely controlled a humanoid robot while making decisions that allowed them to cheat or act generously. We found that:
❌ Teleoperation did not reliably increase cheating
✅ Perticipants were more generous if they teleoperated the robot from the first than third person perspective

We manipulated the visual perspective through which people controlled the robot: through its eyes (first person) or from behind its back (third person)

In this study, in collaboration with Social and Cognitive Neurosciences AgliotiLAB, we examined whether non-immersive, screen-based robot teleoperation affects moral behavior — specifically cheating for personal gain and generosity toward others.

New paper out in iScience 🚀

Does controlling a robot change how honest we are?

Read here 👉 www.cell.com/iscience/ful...

w/ @mmwozniak.bsky.social, Marina Scattolin, Luca Provenzano, Davide De Tommaso, Salvatore Aglioti, @awykowska.bsky.social @iitalk.bsky.social

🌟 It’s International Day of Women and Girls in Science! On this occasion, our PI Prof. @awykowska.bsky.social Wykowska was interviewed by the European University Association about her journey in academia and the challenges and rewards of being a scientist.

Read here 👉 www.eua.eu/our-work/exp...

Across conditions, implicit measures of joint agency and neural markers of action and outcomes were comparable, indicating that joint agency emerged similarly with human and robot partners.

Our findings suggest that sensorimotor cues alone may be sufficient to elicit joint agency with robots.

Participants performed a joint task with another human and with a humanoid robot endowed with a human sensorimotor pattern. They were told either that the robot was controlled by a human or that it was pre-programmed.

In this study, we examined whether human-like sensorimotor coordination or contextual beliefs about intentionality are key for joint agency, using behavioral measures and EEG.

New paper out in Neuropsychologia!

Is human-like sensorimotor coordination enough to elicit joint agency with a humanoid robot?

👉 www.sciencedirect.com/science/arti...

w/ Uma Navare, Veerle Hobbelink, Francesca Ciardo, Kyveli Kompatsiari, Davide De Tommaso & @awykowska.bsky.social

Our results showed that participants' perceptions remained stable: intentionality ratings were consistently high, and value alignment was minimally influenced. These findings suggest that moral attributions to robots might be shaped more by the observer's own values than by brief exposure or framing

Participants interacted with iCub under three different framing conditions:
🔹 No prior information
🔹 A verbal description
🔹 A video showing iCub reacting to moral decision-making scenes

🤖 New paper out in Scientific Reports!

How do people perceive the moral values and intentionality of humanoid robots?

Read here: www.nature.com/articles/s41...

W/ @serehal.bsky.social, Kinga Ciupinska, Davide De Tommaso, @awykowska.bsky.social
@iitalk.bsky.social

w/ Davide Ghiglino, Federica Floris, Davide De Tommaso & @awykowska.bsky.social

@iitalk.bsky.social

Our findings suggest that affordable, toy-like robots may be effective tools to support ToM-related outcomes in clinical settings, highlighting the potential of integrating social robotics into real-world therapeutic interventions.

Using a two-period crossover design, 14 children with ASD received ToM training delivered either through robot-assisted sessions or standard therapy. Robot-assisted sessions led to greater improvement on both ToM and emotion recognition subscales (NEPSY-II) compared to standard therapy.

🤖 New paper out in Frontiers in Psychology!

Can a toy robot support Theory of Mind (ToM) training for children with autism spectrum disorder (ASD)?

Read more here:

www.frontiersin.org/journals/psy...

w/ Caterina Foglino & @awykowska.bsky.social

@iitalk.bsky.social

Across two gaze-cueing experiments, we found:
- Uninstructed gaze cues did not enhance memory
- Task-relevant cues improved performance
- The effect was constrained by working memory load

These findings suggest that the benefits of joint attention are not automatic, but context-sensitive

New paper out in Attention, Perception, & Psychophysics!

Can joint attention enhance visual working memory?

In this study, we examined whether gaze-based attentional cues support memory automatically, or only when they are reliable and task-relevant.

Read here: link.springer.com/article/10.3...

w/ @ilkayari.bsky.social, Davide De Tommaso, @awykowska.bsky.social

@iitalk.bsky.social

A control experiment using neutral color outcomes revealed no such effect, suggesting that emotional content plays a specific role in modulating agency - beyond outcome predictability alone.

We found that the sense of agency was reduced when participants unintentionally caused a sad expression, compared to when they triggered the same expression intentionally.

Across two experiments, participants performed actions intended to elicit a happy expression from iCub. In some trials, the robot unexpectedly displayed a sad expression instead.

🤖 New paper out in Scientific Reports!

What happens to our sense of agency when we unintentionally cause a sad expression in another agent?

In this study, we investigated how people experience agency when their actions result in unintended emotional expressions in a humanoid robot (iCub).

w/ Cecilia Roselli, Uma Prashant Navare, Francesca Ciardo, Davide De Tommaso, Sonja A. Kotz, @awykowska.bsky.social @iitalk.bsky.social

Interestingly, while individual SoA was linked to early neural markers (N1), vicarious sense of agency relied on later cognitive processes indexed by P2 - suggesting different underlying mechanisms despite similar behavioral outcomes.

In this study, we investigated the minimal conditions for vicarious sense of agency toward artificial agents. Across two experiments using intentional binding and EEG, we found that access to a proximal action effect, even without motor execution, was sufficient for vicarious SoA to emerge.

🧠 New paper out in Cognition!

Can we feel a sense of agency over the actions of an artificial agent, even when we’re not acting ourselves?

Read the full paper here: www.sciencedirect.com/science/arti...

@awykowska.bsky.social, head of the IIT @s4hri.bsky.social unit and coordinator of the IIT Center for Human Technologies, is this year’s winner of the André Mischke Prize.

#HumanRobotInteraction #sciencecommunication #policyengagement

yacadeuro.org/andre-mischk...