Blogs
Age of generative AI, misinformation: When seeing is no longer believing
AI-generated images, videos and fake accounts are making deception easier, writes Anthea Mascot Turwomwe
Anthea Mascot Turwomwe. (File)
By: Admin ., Journalist @New Vision
____________
OPINION
By Anthea Mascot Turwomwe
Artificial intelligence (AI) has given humanity extraordinary tools. We can create images, videos, voices and stories within minutes. However, the same technology can also manufacture realities.
After the death and funeral of Oyo Nyimba Kabamba Iguru Rukidi IV, social media was flooded with images, videos and claims about the former Tooro king. Some looked so convincing that people shared them as eyewitness accounts. Alongside genuine information came fabricated material, including manufactured children, pseudo social media accounts and images presented as reality.
This is where the problem becomes bigger than social media gossip. AI has dramatically lowered the cost of deception.
A person no longer needs a television studio, a printing press or a large organisation to manufacture a convincing narrative. A smartphone, a generative AI chatbot and a social media account can be enough.
Even the language of credibility has changed. On X, for example, the blue checkmark is now primarily associated with a paid subscription; it is not, by itself, proof that an account is notable, authentic or an authority on a subject.
More troubling is the deliberate construction of stories to give people what they are anticipating or want to hear. We are seeing the possibility of entire narratives being built around public figures simply because an audience is emotionally prepared to believe them. And the danger extends beyond celebrities.
AI-generated videos are increasingly being used in investment scams, including fake endorsements and fabricated appearances by prominent people. Globally, deepfake-enabled impersonation is already damaging reputations and facilitating financial fraud.
So, should governments regulate AI? Yes, but intelligently.
Even the language of credibility has changed. On X, for example, the blue checkmark is now primarily associated with a paid subscription; it is not, by itself, proof that an account is notable, authentic or an authority on a subject.
More troubling is the deliberate construction of stories to give people what they are anticipating or want to hear. We are seeing the possibility of entire narratives being built around public figures simply because an audience is emotionally prepared to believe them. And the danger extends beyond celebrities.
AI-generated videos are increasingly being used in investment scams, including fake endorsements and fabricated appearances by prominent people. Globally, deepfake-enabled impersonation is already damaging reputations and facilitating financial fraud.
So, should governments regulate AI? Yes, but intelligently.
Regulation should not mean regulating people’s thoughts or giving authorities the power to decide what citizens may believe. It should focus on the technology and conduct: mandatory disclosure of deepfakes, stronger penalties for impersonation and fraud. Regulations should also tackle rapid takedown mechanisms, traceability of synthetic political and public-interest content and greater accountability for platforms that knowingly allow deceptive networks to operate.
The EU has already moved in this direction. From August 2026, its AI Act transparency rules require certain digitally generated or manipulated content, including deepfakes and AI-generated public-interest text without human editorial control, to be disclosed.
Uganda also needs to examine its framework. This is particularly important because Uganda’s Constitutional Court nullified several provisions of the Computer Misuse Amendment Act in March on procedural grounds. This means the country cannot simply assume that existing provisions sufficiently address the new AI environment.
But legislation alone will not save us. The first line of defence is the citizen. Before sharing, ask: Who published this? Where did it originate? Is there an official source? Are credible media organisations reporting it? Is the image old or new? Does the video show signs of manipulation? Am I sharing it because it is true, or because I want it to be true? We must teach our children that digital literacy is as important as reading and writing.
We must teach ourselves that virality is not verification, a blue tick is not truth, a photograph is not necessarily evidence, and a video is no longer automatically proof.
If we fail to build a culture of verification alongside the technology, we may eventually reach a point where society can no longer agree on what actually happened. And when a nation loses its shared relationship with reality, the consequences are far greater than a misleading Facebook post.
The writer is the treasurer of the Public Relations Association of Uganda
The EU has already moved in this direction. From August 2026, its AI Act transparency rules require certain digitally generated or manipulated content, including deepfakes and AI-generated public-interest text without human editorial control, to be disclosed.
Uganda also needs to examine its framework. This is particularly important because Uganda’s Constitutional Court nullified several provisions of the Computer Misuse Amendment Act in March on procedural grounds. This means the country cannot simply assume that existing provisions sufficiently address the new AI environment.
But legislation alone will not save us. The first line of defence is the citizen. Before sharing, ask: Who published this? Where did it originate? Is there an official source? Are credible media organisations reporting it? Is the image old or new? Does the video show signs of manipulation? Am I sharing it because it is true, or because I want it to be true? We must teach our children that digital literacy is as important as reading and writing.
We must teach ourselves that virality is not verification, a blue tick is not truth, a photograph is not necessarily evidence, and a video is no longer automatically proof.
If we fail to build a culture of verification alongside the technology, we may eventually reach a point where society can no longer agree on what actually happened. And when a nation loses its shared relationship with reality, the consequences are far greater than a misleading Facebook post.
The writer is the treasurer of the Public Relations Association of Uganda