By: Quadri Yahya
The dragging that follows when an aide to Nigeria’s president accused the media of “misrepresentations of facts” over spotted watermark of Artificial Intelligence in an image buried the question of whether an original picture “enhanced” with an AI should carry watermark.
A picture reportedly of president Bola Tinubu having lunch with Rwanda’s president Paul Kagame has the watermark of Grok.
When the watermarked image was spotted, many X users concluded the lunch never took place.
A senior special assistant to president Tinubu on media and publicity, Temitope Ajayi issued a statement to clear the air.
“The narrative that the picture of Presidents Bola Tinubu and Paul Kagame taken in Paris yesterday was AI generated is not correct”, the statement with heading “stop press” reads.
“The media report and social media comments that followed are misrepresentations of facts. The picture is real and not AI generated as claimed.
“Both President Kagame and President Tinubu met in Paris [and] had lunch together on Sunday (yesterday). The two leaders later had dinner with President Macron [the] same yesterday evening.
“The picture was taken with a phone and obviously had poor quality. The photographer only later used Grok to improve the picture quality. That is not a reason to conclude it was AI generated.
“The writer or editor should have asked questions before this wrong conclusion”, the statement concluded.
Instead of the statement to clear the air, it further stamped the conclusion that the image is entirely AI-generated.
Watermarked image stir debate on X
“
With all due respect sir, This image is 100% generated by @grok. It is in its generation history (in case you miss the watermark), @masuzafi commented under the statement by the president’s aide.
Worse still, Grok, the AI chat bot on X, said the watermark indicates that the image is AI-generated.
AI misinformation raises concerns
There have been concerns about the use of AI-generated media to spread misinformation.
Initially it was easy for fact-checkers to spot AI-generated media by focusing on properties of picture or video. However, as AI tools advance, identifying what is real vs. AI-generated becomes more difficult.
According to a 2025 study, photorealistic images have blurred the line between authentic and synthetic content, leading to misuse and spread of misinformation.
Thus, researchers have suggested that AI-generated media should be watermarked to allow people to easily detect them instead of relying on what has been called the “passive approach”.
This is because “watermark-based detectors consistently outperform passive detectors, both in the presence and absence of perturbations. Based on these insights, we provide recommendations for detecting AI-generated images, e.g., when both types of detectors are applicable, watermark-based detectors should be the preferred choice,” researchers said.
The real concern
Many questions are open to research around AI use in the media.
For one: The AI watermark on the image of the presidents of Nigeria and Rwanda having lunch caused the confusion.
Watermarking means claiming, and claiming means using AI for assistance takes authorship or creator licence away.
It is clear that content produced by AI is unequivocally AI-generated media whereas, it remains unclear why AI will watermark a picture simply enhanced by it.
Or should AI watermark everything?
