South Africa has multiple pieces of legislation that govern the digital spaces, urgency, adaptability and enforcement are key.
The surge in digitally fabricated content is forcing governments across the world to adapt at a pace unlike any other sector.
While many nations have taken firm approaches toward social media and large language model-based Artificial Intelligence (AI), South Africa has spent over five years redrafting one white paper on digital safety.
The latest flashpoint in the tech freedom debate came last week when X’s in-app AI Grok allowed the generation of thousands of sexualised images from user photographs.
The United Kingdom has moved swiftly to place restrictions on Grok, while Malaysia outright banned Grok earlier this week.
Australia banned social media altogether for under-16s in December, with South Korea having criminalised deceptive AI-generated images in 2024.
Department was ‘act with urgency’
The Department of Communications and Digital Technologies (DCDT) in 2020 published its white paper on Audio and Audiovisual Media Services and Online Content Safety.
The white paper’s most recent redraft was published in July 2025, and comments were extended until September that year.
Additionally, a national AI policy framework was first published in 2024, but varied progress has been made in finalising a way forward.
Chairperson of the communications portfolio committee, Khusela Diko, said the committee had repeatedly called for the department to act with urgency in developing a comprehensive regulatory framework.
Diko stated that AI embedded within social media platforms formed part of the broader Over-the-Top (OTT) and online content services environment.
She considered the dangers of an unregulated digital space to include exploitation, the erosion of human rights, and the spread of disinformation that posed a threat to social cohesion.
“These frameworks are necessary to ensure accountability, responsibility, and parity across the sector, while safeguarding the public interest,” Diko told The Citizen.
“Public education and digital literacy are also critical, so that citizens are empowered to navigate online spaces safely and responsibly.”
‘Constitution remains the supreme law’
Tasked with getting the policies over the line is Minister of Communications and Digital Technologies Solly Malatsi.
Malatsi said social media and AI were useful tools that could be “enormously beneficial to society” across a host of industries.
However, Malatsi said anyone misusing AI to harm others should face the full might of the law and that generating abusive images was “nothing short of disgraceful”.
He said South Africa’s constitution was world-renowned for its emphasis on proportionality.
“This concept means that one right does not always trump another right. A special test must be applied to decide whether, for example, the right of freedom of expression should be upheld over the right to dignity,” Malatsi told The Citizen.
“There is no one-size-fits-all rule or law. The Constitution remains the supreme law of South Africa and therefore the benchmark for rights and freedoms.”
Malatsi explained that the country valued its “hard-won freedoms” and that legislation should not spoil those freedoms or allow nefarious actors to exploit technology’s capabilities.
“Like anything else, however, people and businesses can use both social media and AI to promote their own agendas and can intentionally abuse these platforms and technologies,” he said.
Like regulating a tool
Diko and Malatsi both cited the multitude of existing legislation — including the Cybercrimes Act — that protects citizens and punishes transgressors.
But regulating platforms is trickier.
Founder of The LITT Institute, Nerushka Bowan, used an analogy about utility to illustrate the difficulty in regulating AI tools.
“Regulating the capabilities and features themselves is not as practical. That would be akin to trying to regulate a kitchen knife,” Bowan told The Citizen.
“One person can use the knife to cook a gourmet meal, and another can use the same knife to stab their neighbour.
“The knife remains exactly the same tool. It would be impractical to have a regulation that says ‘don’t use your kitchen knife to stab your neighbour’, she explained.
Bowan also cited the enforcement of existing laws, while suggesting the current legislation could be expanded to cover new tech.
“Oftentimes it is the enforcement of these laws that we could further improve on. The difficulty there could be to unmask the perpetrators behind these virtual crimes, and the ability of these criminals to be based anywhere in the world,” she said.
‘The duty of care’
Bowan said she agreed with Australia’s approach to protecting young people and would welcome such a move in South Africa. She is also eager to see the publication of South Africa’s AI policy.
“In addition to the prosecution of crimes, having regulatory guidelines for users and platforms can further assist in creating a safer digital landscape.
“It is a fine line between over-policing, while also preserving each individual’s right to freedom of expression — a right enshrined in our South African Bill of Rights,” Bowan concluded.
Diko said the committee wished to see close cooperation between government, regulators, industry players, and civil society, while also considering South Africa’s unique needs.
“These realities include persistent digital inequalities, high unemployment, linguistic and cultural diversity and the protection of indigenous knowledge systems,” Diko explained.
Malatsi remained steadfast in his commitment to the Constitution and foundational principles.
“The protection of citizens, including their right to privacy and dignity, is sacrosanct. So is the right to freedom of expression.
“All of these require is a legal balancing act. The proportionality test is not that different from the duty of care,” the minister concluded.
NOW READ: Is AI slop destroying trust online?
