The rise of conversational agents (CA) in the media is increasingly changing our news consumption activity. Whether asking CA for the news highlights or consulting it for a specific topic, CA plays a role in the opinion formation process. This project aims to understand the effects of CA on different dimensions of opinion. The emerging communication phenomenon points to two important shifts in the news environment. On the one hand, CA can accelerate the role of algorithms as content moderators.

The role of news moderation is usually associated with human decision-making, and this new role given to algorithms can limit information diversity, challenging the requirements for a healthy democracy. On the other hand, conversational agents can also serve as trusted systems for providing consistent information, reducing content diversity compared to other news sources such as social media. In both cases, CA can have a role in information distribution, influencing the opinion formation process, also known as opinion power.

A topic that has been unexplored by current research. Thus, this project asks how conversational agents affect opinion, bridging the gap on the role of CA in opinion formation. The project will address three key questions through three different studies:

1. How do conversational agents influence opinion salience?
2. How do conversational agents influence opinion certainty?
3. How do conversational agents influence opinion ambivalence?

At the end of the project, the studies will enlighten from a multidisciplinary perspective the opinion power role of CA on the individual.