- AI deepfakes are creating a new corporate trust challenge, making fake communications increasingly convincing.
- South African regulators are warning of deepfake investment scams using trusted public figures and brands.
- Digital provenance helps verify authentic content through secure metadata that records its origin and edits.
- Content Credentials and the C2PA standard are gaining global adoption across technology companies and camera manufacturers.
- Verifiable authenticity is becoming a business imperative, not just a communications responsibility. *
For years, organisations have invested heavily in protecting themselves against cyberattacks and online misinformation. They are now confronting a different challenge. The question is no longer simply whether false information will circulate, but whether audiences can distinguish authentic corporate communications from increasingly sophisticated fabrications.
A chief executive appears to endorse an investment platform. A central bank governor seems to announce market-sensitive information. A finance director invites investors to join a WhatsApp group promising exceptional returns.
The video is convincing, the voice sounds authentic and the branding appears legitimate. None of it happened.
South African businesses are already dealing with the consequences. The South African Reserve Bank has warned the public about AI-generated videos that falsely portray Governor Lesetja Kganyago involved in televised confrontations with media personalities, and fake content promoting fraudulent investment schemes.
Deepfake videos
The Financial Sector Conduct Authority has repeatedly warned that fraudsters are increasingly using deepfake videos and the identities of well-known public figures – including Springbok captain Siya Kolisi, President Cyril Ramaphosa and businessman Patrice Motsepe – to lend credibility to fraudulent investment schemes.
These incidents reflect a broader pattern. Earlier this year, Hong Kong police disclosed that a multinational company had lost more than $25-million after employees were deceived by AI-generated video and audio impersonating senior executives during a video conference.
The World Economic Forum has identified misinformation and disinformation enabled by artificial intelligence as one of the most significant short-term global risks because of its capacity to undermine confidence in institutions, markets and democratic processes.
Proving digital provenance
The challenge extends beyond misinformation.
Businesses have always had to respond to false claims and online rumours. What has changed is the ease with which convincing digital content can now be produced.
Images, voice recordings, videos and documents that once required specialist skills and significant resources can now be generated within minutes, using widely available tools. The technical barrier to creating credible deception has fallen sharply.
That has important implications for reputation management.
A fabricated earnings announcement can influence markets before an organisation has time to respond. A manipulated executive statement can spread across multiple platforms before journalists or investors establish that it is false. By the time a correction is issued, the original content may already have reached millions of people.
For decades, organisations have approached trust largely as a communications challenge. Editorial standards, governance processes and rapid crisis response remain essential, but they address only part of the problem. Businesses must now consider how they will demonstrate that legitimate content originated from them and has not been altered.
Transparent record
The concept of digital provenance offers one possible answer.
In the art market, provenance refers to the documented history of an object. Before purchasing a valuable painting, collectors want evidence of its origin, ownership and restoration history because that record establishes authenticity. The same principle is increasingly being applied to digital content.
Rather than relying solely on software that detects manipulated media after publication, digital provenance allows creators to attach verifiable information to content when it is produced. The objective is not to guarantee that every image or video is truthful, but to provide a transparent record showing where it originated, what tools were used to create it and whether it has subsequently been edited.
The best-known initiative in this field is the Coalition for Content Provenance and Authenticity (C2PA), an open technical standard supported by technology companies, publishers and camera manufacturers.
Through systems such as Content Credentials, creators can attach cryptographically signed metadata to digital files, recording information including when the content was created, which device or software produced it, and whether generative AI was part of the production process.
Recording authenticity
Support for the standard is expanding. Adobe has integrated Content Credentials across its creative applications. Microsoft and Google have joined industry efforts to strengthen provenance standards, while OpenAI has introduced Content Credentials for images generated through ChatGPT.
Camera manufacturers such as Canon, Nikon, Leica and Sony are also developing hardware capable of recording authenticity information at the point an image is captured.
The technology is not a complete solution, because metadata can be removed under certain circumstances and provenance information is valuable only if software platforms preserve it.
The organisations behind these standards acknowledge that provenance must operate alongside watermarking, detection technologies, platform governance and stronger public understanding of synthetic media.
An operational issue
Even with the limitations in mind, the discussion about trust is changing. Organisations have traditionally concentrated on ensuring that what they publish is accurate. Increasingly, they will also need systems that allow others to verify that the content genuinely originated from them.
Authenticity is becoming an operational issue, rather than simply a communications concern, says Flow Communications chief technology officer Richard Frank.
“We’re reaching a point where proving the origin of content is becoming part of maintaining trust. Clients still want people to share their content, but they also want audiences to recognise that it is authentic.”
Convincing fabrication is becoming commonplace on the internet. This means, in future, credibility will depend less on how persuasive content appears than on whether organisations can demonstrate where it came from and how it was created.
For businesses, the next competitive advantage may not lie in producing more content, but in producing content that can be verified as authentic.
* Summary created by AI.
Thelma Ngoma-Mavhunga is senior communications specialist at Flow Communications.
