I have put on my ICT hat to analyse the legal and technological complexities surrounding the ongoing case involving controversial businessman Wicknell Chivayo, his ex-wife Sonja Madzikanda and her mother Tabitha Madzikanda.
This is not merely a high-profile domestic fallout involving allegations of cyberbullying. It is a case that could test the State’s ability to investigate, authenticate and successfully prosecute offences involving Artificial Intelligence-generated content.
The prosecution alleges that AI-generated photographs and videos were circulated suggesting that Chivayo had met South African President Cyril Ramaphosa ahead of his official visit to Zimbabwe. The State is reportedly expected to rely on these digital images and videos as key evidence.
That is where the real challenge begins.
The problem of proving authenticity
In traditional criminal investigations, evidence such as fingerprints, handwriting, CCTV footage and call records can usually be subjected to forensic verification. AI-generated content changes that landscape entirely.
- Power crisis needs practical solutions
- Dangers of Arab Spring in SA
- Zimbabweans fork out US$24m for SA permits
- Power crisis needs practical solutions
Keep Reading
Modern generative AI tools can now create highly realistic videos, cloned voices, fabricated photographs and manipulated conversations that are increasingly difficult to distinguish from authentic material. Some technologies have become so advanced that even trained ICT professionals may struggle to immediately detect manipulation without specialised forensic tools.
For the State to secure a conviction, it will not be enough to merely present videos or images in court. Prosecutors would need to prove:
* Who created the content
* Which device was used
* Which software generated it
* Whether the accused knowingly distributed it
* Whether the material was maliciously fabricated
* Whether metadata supports the allegations
* Whether the chain of custody remained intact
Without that, the defence could argue that the evidence was fabricated, altered, planted or generated by unknown third parties.
The deepfake era has arrived
Zimbabwe’s justice system is now entering the dangerous era of deepfakes.
A deepfake is AI-generated or AI-manipulated media designed to imitate real people. Across the world, governments, intelligence agencies and courts are struggling to adapt because the technology is evolving faster than legislation.
This matter could become one of Zimbabwe’s first major tests of whether local legal and forensic systems are prepared for AI-assisted cybercrime.
The State would likely require:
* Digital forensic experts
* AI detection specialists
* Metadata analysts
* Cyber investigators
* Server and cloud data tracing
* Social media forensic reports
The challenge, however, is that Zimbabwe still has limited institutional capacity in advanced cyber forensics compared to more developed jurisdictions.
Chain of custody will be critical
One of the biggest weaknesses in digital investigations is chain of custody.
The State must demonstrate precisely:
- Where the evidence originated
- Who extracted it
- Which devices stored it
- Whether it was modified
- Who had access to it
A single unexplained alteration can render digital evidence unreliable.
If screenshots or media files were merely forwarded through platforms such as WhatsApp, the defence may challenge whether the original files even exist in authentic form.
AI evidence can easily create reasonable doubt
The burden of proof in criminal law remains proof “beyond reasonable doubt”.
That standard becomes significantly harder to satisfy when AI is involved.
Defence lawyers could argue that:
* The content was generated by someone else
* The accused were framed
* The media was edited after distribution
* Accounts were hacked
* Devices were compromised
* Images were synthetically created elsewhere
In ICT investigations, reasonable doubt expands rapidly when forensic evidence is incomplete.
Legal gaps may soon emerge
Zimbabwe’s cyber laws were largely drafted before the explosive rise of generative AI. While cyberbullying provisions exist, the legislation may not fully anticipate the sophistication of AI-generated media.
This case could expose major legal gaps regarding:
* Admissibility of AI-generated evidence
* Verification standards
* Digital authentication procedures
* Expert witness requirements
* Liability for reshared content
* Responsibility for synthetic media
Courts may soon require entirely new evidentiary standards for AI-related offences.
The danger of politically sensitive digital content
The alleged involvement of images linked to President Ramaphosa elevates the sensitivity of the matter.
Once political figures become associated with AI-generated content, cases move beyond private disputes into issues involving diplomacy, reputational harm, misinformation and national image management.
That increases pressure on investigators to establish authenticity with precision and professionalism.
From an ICT perspective, this is not simply a celebrity court case. It is a warning shot.
Artificial Intelligence is now capable of manufacturing believable realities. Zimbabwean institutions — courts, police, prosecutors, banks, companies and ordinary citizens — are entering an era where seeing is no longer believing.
The Madzikanda versus Chivayo matter may ultimately become one of Zimbabwe’s first major legal battles in which the courts must answer a frightening modern question:
Can digital evidence still be trusted in the age of Artificial Intelligence?
