Google's SynthID technology succeeded in detecting a fake image created by artificial intelligence, in one of the most prominent cases where the technology proved its effectiveness in combating 'deepfake' content.

In recent days, an image circulated on X and Reddit platforms that purportedly showed US Senator Mitch McConnell lying on a bed in a hospital, surrounded by medical tubes and appearing in critical health condition, sparking a wave of speculation about his health status.

But the famous fact-checking site Snopes confirmed on Wednesday that the image was entirely fake, explaining that an examination revealed it contained the SynthID digital watermark that Google adds to AI-generated images, quickly exposing the image's true nature, according to a report by TechCrunch and reviewed by Al Arabiya Business.

How did Google expose the forgery?

The SynthID technology relies on adding an invisible digital watermark within the image, such that the average user does not notice it, but it remains detectable by dedicated tools.

This watermark is characterized by remaining embedded in the image even after it is reposted or screenshots are taken across different platforms, which is what happened with the fake image attributed to the US senator.

This incident is seen as a practical success for anti-deepfake technologies, after they managed to expose an image that had spread widely before it could become accepted as fact.

Rumors about McConnell's health

The spread of the fake image came amid rising rumors about the health condition of Senator Mitch McConnell, after he was hospitalized following an emergency call on June 14, before largely disappearing from public view, opening the door to a wave of speculation about the deterioration of his health.

However, the circulated image was nothing more than AI-generated content with no connection to reality.

What is SynthID technology?

Google launched SynthID technology at the Google I/O 2025 developer conference, aiming to provide a reliable means of identifying images produced by artificial intelligence.

The technology works by embedding an invisible digital signature within the image, which can be detected later without affecting the image quality or appearance.

However, this technology only works if the image generation model itself participates in the program and adds the digital watermark when creating the image.

Since its launch, Google's Gemini models have supported SynthID by default, while OpenAI joined the program in May 2026 as part of its efforts to curb the misuse of image generation tools.

In contrast, Anthropic is not currently participating in the program.

Users can verify the presence of the digital watermark in images by asking one of the Gemini models or by uploading the image to OpenAI's public tool dedicated to checking AI-generated images.

Advertising material

Advertising material