AI algorithms and the new power of public opinion

In this digital era, public opinion is formed in newspaper columns, television studios, political meetings and on social media platforms, where millions of people react to an event within seconds. Behind this enormous flow of public expression are artificial intelligence systems and algorithms that decide, often invisibly, what people see, discuss and which voices become dominant.

This largely unregulated and fast changing environment makes me wonder: what happens when digital crowds begin to act as judges?

The rise of social media has given ordinary citizens unprecedented power. A single post attracts thousands of comments, while a video travels across borders before traditional media organisations have even confirmed the facts. People such as the late Prof Jason Ardey, whose public profile and actions became the subject of online discussions that resulted in the end of his life, are evidence of the dangers of digital visibility.

Algorithms are designed to keep users engaged. They analyse behaviour, including what people click on, share, comment on and watch. Content that generates strong reactions is often given greater visibility. Unfortunately, controversy, anger and outrage always attract more attention than balanced explanations.

We have seen accusations spread faster than evidence.  People have been judged by thousands of strangers who have never met them and know little about the circumstances surrounding their particular incidents. Once a narrative becomes popular, correcting it can be extremely difficult, hence online content turned into digital courtrooms.

Public discussion has always been part of society. The deeper concern is that algorithms are amplifying particular opinions at enormous speed and scale, creating the impression that the loudest voices represent the majority.

Digital crowds easily become powerful forces. They support persons to destroy reputations, influence companies, or even pressure institutions into taking action. At ibest, this power can expose wrongdoing and give marginalised communities a voice. At worst, this power can encourage harassment, misinformation, online punishment and even death without due process.

Artificial intelligence adds another layer to the problem. AI systems have the power to identify trends, recommend content, generate summaries and produce convincing text, images and videos. These technologies help people understand complicated subjects, but they also make misinformation easier to create and distribute.

For public figures and private citizens alike, reputation is now being shaped by an unpredictable combination of human emotion and machine recommendation. Someone wakes up to discover that their name is trending, without understanding why. Within hours, thousands or millions of people have formed opinions about them based on fragments of information.

This creates a challenge for journalism as well. Professional journalists are expected to verify information, identify sources and provide context. Social media operates according to very different incentives. Speed often wins over accuracy, while emotional content outperforms careful reporting.

I am not suggesting that we silence digital crowds. Freedom of expression remains essential. Nor should algorithms automatically be treated as enemies. They help people discover information, connect communities and make sense of enormous quantities of data.

The real challenge is accountability. Technology companies need greater transparency about how recommendation systems work and how harmful or demonstrably false information is amplified. Users also need stronger digital literacy so that they can distinguish evidence from speculation and reporting from viral commentary.

Most importantly, society must resist the temptation to confuse popularity with truth.

If thousands of people believe something, that does not automatically make it true. If an algorithm repeatedly places a story in front of us, that does not mean the story is important or accurate. And if an online crowd condemns someone, that does not necessarily mean justice has been served.

The story surrounding the late Prof Ardey, viewed within this wider digital landscape, illustrates a much larger issue facing modern society, which is that who gets to decide the truth after technology has amplified millions of individual judgments?

A healthy society requires evidence, accountability, context and the willingness to reconsider an initial judgment. Digital platforms have given people extraordinary power, but with that power comes responsibility.

The internet has created a world where everyone is now a commentator and every controversy is a public trial. The task now is to ensure that the speed of digital judgment does not outrun the slower, but essential, processes of truth and justice.

Toshi Bwana is the Founding Trustee of Umoja Conservation Trust (UCT).