Electoral lies and political speech, the quest for truth - IOL

SOUTH Africa is right to confront deliberate electoral disinformation. But the legitimacy of that intervention will depend on maintaining a disciplined boundary between deception that threatens an election and the exaggeration, opinion, criticism and uncomfortable disagreement that democracy must tolerate.

The most dangerous lie in an election may not be about a politician. It may be about the election itself.

A message says a polling station has moved. A convincing video appears to show an electoral official admitting that ballot papers have been manipulated. A candidate seems, in a recording, to say something never actually said. By the time the fabrication is challenged, thousands of people may have seen it, shared it, and acted on it.

That is the information environment in which South Africans will vote in local government elections on November 4. 

The Electoral Commission of South Africa recently introduced a new Disinformation Code intended to protect the integrity of the election. Political parties and candidates must verify information before distributing it, publicly correct false or misleading material within 36 hours of becoming aware of it, avoid fake accounts and bot networks, label AI-generated content, and make political advertising identifiable.

The old democratic assumption that falsehood can simply be defeated by better information belongs to a slower world.

A convincing fabrication no longer requires a sophisticated political machine. It requires a phone, an audience and an algorithm willing to reward whatever captures attention.

Artificial intelligence has made it cheaper to create believable voices, images and video. Social media can distribute them before a journalist has verified them, an opponent has answered them, or an electoral authority has begun considering a complaint.

Some falsehoods strike directly at the ability to vote. A fabricated announcement that an election has been postponed, that a polling station has moved, or that ballot papers have been compromised does something more serious than distort political debate. It interferes with the conditions under which citizens exercise political rights.

Doing nothing is therefore not necessarily a principled defence of free expression. Sometimes it simply leaves the information space to whoever can manufacture confusion fastest.

But that cannot be where the argument ends.

Governments describe their records generously. Opposition parties describe the same records rather differently. Candidates exaggerate, simplify and accuse. Statistics are selected. Promises are stretched. One person's dishonesty can be another person's interpretation.

Democracy has always required a tolerance for this disorder.

South African law already recognises that electoral freedom does not mean a licence to manufacture false information. But the Constitutional Court has also been wary of allowing those restrictions to suffocate political expression.

That caution matters even more now.

In a case decided in 2015 arising from the Nkandla controversy during the 2014 election campaign, the Concourt had to consider whether a provocative campaign message constituted prohibited false information or protected political opinion. The dispute forced the Court to confront the uncomfortable nature of democratic speech: partisan, argumentative and sometimes abrasive.

The court’s majority held that the message was an expression of comment or opinion rather than “false information” under the Electoral Act.

The principle behind that judgment should not be discarded simply because the technology has become more dangerous.

There will be easy cases under the new code. A fake account impersonating an electoral official should attract intervention. So should an intentionally fabricated announcement about voting procedures, a synthetic recording deliberately presented as genuine, or an organised network designed to deceive voters about the electoral process.

The more important test will come at the edges.

What happens when a party uses accurate figures but presents them in a misleading way? When a candidate makes an allegation based on disputed evidence? When an activist shares information later shown to be wrong but genuinely believed it to be true? When satire is obvious to some people and entirely convincing to others?

The danger is allowing the category of “disinformation” to expand until it becomes shorthand for speech that is disputed, embarrassing, politically inconvenient, or simply incorrect.

Once that happens, the regulatory process itself can become another political weapon.

Why answer an opponent if you can report them? Why rebut criticism if you can call it disinformation?

That would also place electoral authorities in an increasingly difficult position.

Institutions responsible for administering elections must protect the process from deliberate interference while remaining open to vigorous criticism of their own conduct.

Confidence in an electoral commission cannot be secured by making criticism of it softer. Trust is earned when institutions can withstand scrutiny and demonstrate, publicly and convincingly, why false accusations are false.

The answer must therefore be precision.

Where information is capable of objective verification, deliberately fabricated and likely to interfere with electoral participation or integrity, intervention can be firm.

Where the dispute concerns political interpretation, inference, rhetoric or opinion, the threshold should be considerably higher.

When material is ordered to be corrected or political actors face consequences, voters should be able to understand why. Decisions should be reasoned, transparent and capable of independent scrutiny.

Simply attaching the label ‘false’ to contested information may remove a piece of content. It will not necessarily remove the suspicion that allowed it to flourish.

No electoral code can make citizens pause before forwarding something that confirms what they already believe. No regulator can enter every private messaging group before a fabricated claim begins to spread.

Technology companies cannot design platforms that reward outrage and virality and then behave as though the political consequences belong entirely to governments and voters.

Political parties cannot exploit synthetic media, anonymous networks and manipulated content when useful, and invoke freedom of expression when challenged.

Democratic participation requires more than arriving at a polling station. It requires judgement: the willingness to question information we desperately want to believe, particularly when it confirms everything we already think about those we distrust.

The choice is therefore not between regulation and freedom.

Democracies need institutions capable of acting against deliberate electoral deception. They also need citizens and institutions capable of living with disagreement.

South Africa is trying to find the difficult space between those two demands.

Other democracies should watch carefully. Deepfakes will improve. Synthetic voices will become harder to identify. Artificial networks will become more convincing. Pressure on electoral authorities to decide what is true, false and dangerous, will grow.

Some of that power will be necessary.

A democracy must make deliberate electoral deception difficult.

It must make political disagreement easy.

Its health may depend on preserving both principles at the same time.

No electoral code can make citizens pause before forwarding something that confirms what they already believe. No regulator can enter every private messaging group before a fabricated claim begins to spread.

Technology companies cannot design platforms that reward outrage and virality and then behave as though the political consequences belong entirely to governments and voters.

Political parties cannot exploit synthetic media, anonymous networks and manipulated content when useful, and invoke freedom of expression when challenged.

Democratic participation requires more than arriving at a polling station. It requires judgement: the willingness to question information we desperately want to believe, particularly when it confirms everything we already think about those we distrust.

The choice is therefore not between regulation and freedom.

Democracies need institutions capable of acting against deliberate electoral deception. They also need citizens and institutions capable of living with disagreement.

South Africa is trying to find the difficult space between those two demands.

Other democracies should watch carefully. Deepfakes will improve. Synthetic voices will become harder to identify. Artificial networks will become more convincing. Pressure on electoral authorities to decide what is true, false and dangerous, will grow.

Some of that power will be necessary.

A democracy must make deliberate electoral deception difficult.

It must make political disagreement easy.

Its health may depend on preserving both principles at the same time.

Neetu Chetty is a legal academic, higher-education leader, social justice activist and mother whose work spans law, education, governance, leadership and technology. With BProc, LLB and LLM degrees from the University of KwaZulu-Natal, she is also an ETDP-certified Skills Development Facilitator and has spent much of her career challenging how we educate, lead and create opportunity for others.

** The views expressed do not necessarily reflect the views of IOL or post.co.za