A hateful comment appears beneath a public post by an LGBT person. Then another. Soon the comments multiply, along with threats and false claims. The person can block the accounts spreading hate, report them to the platform, or leave social media altogether. But the next time they consider speaking publicly, they know what to expect.

This is how online hate affects public debate. Sometimes there is no need to persuade an opponent with arguments. It is enough to make them unwilling to speak.

Through my work in human rights, I see how hate in public spaces affects LGBT people. When someone shares their experience publicly, attention often shifts away from what they said and towards their identity, or even their right to be visible at all. Women, members of ethnic and religious minorities, journalists and human rights activists face similar pressure. The reasons for the attacks differ, but the result is often the same: fewer voices remain in public debate.

Artificial intelligence is changing the scale of this problem. Today, convincing false text, images and videos can be created quickly. The same hostile narrative can be tailored to different audiences and spread far more efficiently. A person targeted by such content may feel its effects long before anyone establishes who created it.

Technology does not decide how it is used. People build the tools, set platform rules and decide how to respond to reports of harm. We therefore need to talk about responsibility: what do content creators, platforms and institutions do when they see hate spreading ever faster in the spaces they shape or oversee?

The advice to “just stop reading the comments” does not solve the problem. When someone faces threats or lies, or their professional reputation is unjustly damaged, they need a clear way to seek help and to understand what will happen next. Platforms must act promptly and explain their decisions to users. Those who experience online hate must also be heard when decisions are made about the protections they need.

The European Union’s Digital Services Act and AI Act provide a framework for this work. Their impact, however, will be measured in practice. Can people report harmful content easily? Will their reports be taken seriously? Can platforms and institutions respond when hate spreads at such speed?

A public space cannot be considered open if taking part in it consistently leaves some people feeling unsafe. To protect free debate, we must ensure that the ability to participate does not depend on a person’s capacity to “withstand” hate.

Originally published on LRT.lt: https://www.lrt.lt/naujienos/pozicija/679/3061178/egle-kuktoraite-kas-atsakys-uz-algoritmu-platinama-neapykantos-kalba