The latest buzzword in all aspects of our lives is AI and its ubiquity means everything, including journalism is impacted either in a positive or negative way and it is crucial that those on the polar ends of the news cycle – producer and consumer – keep their eyes and ears peeled as well as their minds open to the opportunities and threats these technologies present.

For the institution of journalism, whose life blood is unbiasedness and credibility, the threats are where the danger is, with an increasing number of content creators making use of AI tools to disseminate their produce put forward at times as breaking news, despite the disregard for journalistic principles.

The motivations behind the creation of this content, which sometimes contains disinformation and information, vary, and according to UKZN’s associate professor for Information Systems and Technology, Professor Upasana Gitanjali Singh, “understanding the motivation behind such content is key to addressing it”.

“In the South African context, disinformation and misinformation are often driven by political, economic, or ideological goals, as well as by malicious actors seeking to exploit social divisions — particularly around issues like race, inequality, and public health.

Also read: The AI assistant that keeps my mind from overheating

“When such content is presented as legitimate news, the intent is often to influence public opinion, discredit institutions, or undermine social cohesion. In some cases, it’s monetarily motivated – fake news sites generate ad revenue by attracting clicks with sensationalist or misleading headlines,” said Singh.

By way of example, the IEC, the judiciary, or the Department of Health were cited by Singh as institutions which could be targeted, with the latter one recently resorting to issuing posts on its social media platforms cautioning about fake news circulated as an anti-vaccine video in which a deepfake of a well-known SABC news anchor was used.

“This is a prime example. This was not just a prank — it was an attack on public trust in credible news sources and public health messaging. Such acts are particularly dangerous in a country like South Africa, where historical mistrust of the State and media still lingers, especially in rural or marginalised communities.

“AI has lowered the barrier to creating convincing disinformation. South Africans are already grappling with high levels of misinformation on platforms like WhatsApp and Facebook, and the introduction of AI-generated fake videos, voices, and articles makes this even more challenging,” said Singh.

The dangers posed by the increased prevalence of such content, cautioned Singh, include:

•          The erosion of trust in media institutions, which is already fragile post-apartheid.
•          Public health threats, as seen during Covid-19 vaccine campaigns.
•          Political instability, especially during elections, if fake news is used to manipulate voter sentiment.
•          Polarisation, where communities become entrenched in opposing views based on false narratives.

“Disinformation, when localised using familiar faces and culturally specific language, becomes even more potent and harder to debunk,” said Singh.

AI “hallucinations” – fabricated facts – pose serious risk for journalists who are cautioned not to rely solely on tools like ChatGPT as these can compromise “journalistic credibility”.

“Biases embedded in AI models may inadvertently skew coverage, especially in sensitive South African contexts like land reform, race and inequality,” said Singh, who further cautioned that relying solely on such tools could lead to a “loss of investigative depth” in the field of journalism.

“In South Africa, where journalism often serves as a watchdog over State capture, corruption, and social justice, the reduction of human oversight in newsrooms could lead to a loss of nuance and accountability,” said Singh.

Also read: Water curtailments to be reintroduced in September

Advisably, AI tools like ChatGPT should be used in newsrooms for assistance with newsmakers ever mindful of the serious risks posed.

These AI tools also raise ethical concerns, which, Singh said, include loss of employment, a question about who would account for errors in AI generated articles, and whether newsrooms should be transparent about AI generated content.

“There’s also a risk of devaluing journalistic labour, where editors begin to see journalists as redundant if AI can produce content quickly – albeit shallow and without context. The devaluation of journalistic labour is already a concern. If machines can generate content in seconds, what incentive remains to invest in deeply researched, human-written stories?” questioned Singh.

There, however, are positives to AI as it can assist newsrooms with content management, audience engagement and analytics.

The negatives, said Singh, include an inundation of the media space by AI-generated content, which adversely impacts the pockets of traditional media houses.

“In South Africa, public broadcasters like the SABC and print media like Independent Media and Media24 are under immense pressure. The influx of free, AI-generated content – much of it misleading – draws attention away from credible, fact-checked journalism undermining revenue models,” said Singh.

News consumers in South Africa should be vigilant. Key tips include:

• Verify the source: Trust known media brands (e.g., Daily Maverick, News24, SABC, Mail & Guardian).
• Check for bylines: AI-generated articles often lack human author names.
• Watch for unnatural language: Overly generic or repetitive language can be a red flag.
• Reverse image search: Use tools like Google Lens to check if an image has been altered or repurposed.
• Use fact-checking platforms: Resources like Africa Check and Real411 are reliable for verifying suspicious claims.
• Be cautious of WhatsApp forwards and TikTok news, which often spread AI-manipulated content quickly.

“Disinformation is rarely random. It’s often driven by political, ideological, or financial motives. In South Africa, with its history of social inequality and polarisation, false narratives are easily weaponised to stir unrest or discredit institutions. And AI tools now make it easier and faster than ever to manufacture such content.

“AI presents both opportunities and risks for journalism. But in South Africa’s complex socio-political landscape, where misinformation can have life-or-death consequences, the stakes are higher than ever. As we move forward, responsible use of AI — combined with ethical oversight, policy reform, and public education — will be critical to protecting the truth,” said Singh.

For more from the Highway Mail, follow us on Facebook , X and Instagram. You can also check out our videos on our YouTube channel or follow us on TikTok.

Click to subscribe to our newsletter here