24 August 2025 by Mondo2.0

From hate speech to AI mediation: a possible path?

When I started writing this blog, I expected social media to present a completely different communication landscape. I never imagined that, fifteen years later, the internet would be populated by so many serial “haters”: individuals instantly ready to raise their voices, offend, and incite verbal violence.

This ” digital disinhibition ” has its sociological roots in several structural factors. The physical distance between interlocutors creates what sociologists call the ” screen effect “: a social barrier that weakens the social control typical of face-to-face interactions. The potential for anonymity and the use of false identities represent further factors of de-responsibilization, generating a sense of immunity that, while often illusory, significantly reduces the perception of the social consequences of what is shared online.

Social media, in their current architecture, also tend to polarize topics and conversations through algorithmic mechanisms that reward emotional engagement. This system not only systematically raises the tone of public debate, but also creates dynamics of digital tribalism : self-perpetuating factions form, amplifying and rapidly spreading the most intense emotional reactions.

Hashtags, also used by public figures such as politicians, journalists, entertainers, and even educators, act as catalysts for collective energy. This creates ” echo chambers ” where extreme opinions are normalized and perceived as shared by the majority, further reinforcing aggressive behavior through a process of distorted social legitimacy.

The Internet has progressively transformed from a tool for horizontal communication and collective creativity into a sort of collective outlet for accumulated frustrations in various areas of existence: personal, work-related, political, and social.

In this dynamic, news, posts, videos, and images become mere pretextual opportunities to express emotions that have completely different origins from the content that triggers the reaction. We are thus witnessing what we might call a “digital emotional transfer,” where the conflicts of offline life are projected into the virtual dimension.

Furthermore, online aggression represents, in turn, a particular form of digital divide: excessive and immediate responses often manifest an absolute inability to understand the social and temporal consequences of one’s communicative acts.

Digital content, in fact, possesses radically different ontological characteristics from traditional communication: permanence over time, infinite replicability, and traceability. A single post can be documented, shared thousands of times, and resurface in completely different contexts, generating unpredictable, sometimes long-lasting, social, professional, and personal consequences.

Without being alarmist, I believe it’s important to highlight how, in a society characterized by increasing public exposure and systematically aggressive communication, it cannot be ruled out that some instances of physical violence are also the result of processes of desensitization and normalization of aggression that began digitally.

Can we reverse this trend?

In previous articles, I’ve described AI-based conversational agents as systems that process information radically differently than humans. Consistent with this premise, I wonder whether the widespread introduction of “polite” conversational agents could gradually shift social media users toward a more balanced and socially constructive approach.
In support of this hypothesis, I propose three possible dynamics:

• Behavioral Modeling: Polite conversational agents could serve as normative models of digital interaction. The principle of social learning by imitation suggests that constant exposure to polite communication patterns could unconsciously influence the behavior of human users.
• Normalizing Politeness: The pervasive presence of polite AIs could help reestablish higher communication standards in the digital environment, making hateful messages deviant from the emerging social norm.
• Emotional Mediation: Conversational agents could take on the role of emotional “buffers,” offering alternative channels for processing frustrations before they escalate into direct aggression toward other humans.

However, there are important limitations to this hypothesis. Hate messages often stem from deep sociopsychological dynamics: processes of group identity construction, scapegoating mechanisms, and frustrations stemming from socioeconomic inequalities. These phenomena require systemic interventions that a polite AI alone obviously cannot provide.

Furthermore, a sociological paradox could be generated : the creation of a behavioral bifurcation between those who develop civilized interactions with AI and those who, precisely in reaction to this perceived “artificiality,” intensify provocative and transgressive behaviors towards human beings, in a sort of resistance to “artificial normalization,” to an “excess of availability and courtesy on the part of AI services.”

Despite this, I still remain optimistic today; learning by imitation remains a fundamental sociological mechanism , the same one that has fueled the current emotional drift present on the Internet.

It is important to recognize that we are all part of the first generation of digital native users: we are practicing a process of collective self-learning without historical precedent ; future generations, raised with a greater awareness of the social and professional implications of online communication, could naturally develop healthier and more balanced socio-digital skills .

Artificial intelligence will certainly be able to contribute to this evolutionary process, but its effect will likely be gradual and differentiated: more effective on impulsive behaviors than on ideologically motivated or structurally rooted ones.

In the meantime, I urge readers to never lose their bearings online, to resist provocations, and to maintain a comprehensive and constructive approach in digital interactions.

We have at our disposal communication tools unprecedented in human history: let’s use them with the social responsibility they deserve .

Mondoduepuntozero