OPINION: Google unveiled its 9th generation of Pixel smartphones this week, with 2024’s line-up including the Pixel 9, Pixel 9 Pro, Pixel 9 Pro XL and Pixel 9 Pro Fold. During a year in which companies are racing to one-up each other with the smartest AI, Gemini stole the show for Pixel fans as Google dedicated plenty of time to how the AI assistant will improve experiences on its latest smartphones.
However, there’s one AI tool that flew under the radar during the bustle of announcements and that is how Google Gemini can now identify AI-created content.
“Pixel helps protect you from spam, scams and other threats, and also looks out for your safety”, wrote a Google spokesperson in press releases for the Pixel 9 and Pixel 9 Pro Fold.
“It can now help you to check the background and context of an image that you come across online, showing information, when available, that can indicate if it was created with AI. Just circle the image on your screen to search it and pull up the ‘About this image’ feature”.
This new ability appears to be an extension of Google’s Circle To Search function, a feature that encourages users to circle content onscreen to learn more about it. Users can circle products to buy them, landmarks to get directions, plants to identify them and even actors to head directly to their IMDb pages and learn where they might recognise them from.
The tool was first announced alongside the Samsung Galaxy S24 in January of this year but has since expanded to include more phones, including Google’s own Pixel range.
Now, in something of a reverse Turing test, users can circle images to learn whether or not they were created using AI.
There are a couple of flaws with this concept from the get-go. Firstly, almost all photos processed on modern smartphones are hit with some form of AI. Pixel phones, for example, are packed with AI editing tools from the Magic Eraser and Best Take to the brand-new Add Me feature that merges two photos to ensure no one is left out of a group shot.
Adobe’s CTO of Digital Media Business Ely Greenfield pointed out something similar regarding the iPhone when I interviewed him about AI art and content flagging earlier this year.
“If we wanted to flag any image where AI was used in the making of the image and put a big warning label on it, every photo that came out of an iPhone would be flagged because AI is used in the rendering, in the capturing of the image. It is not actually the raw luminosity values that the sensor is seeing. Apple actually has AI, wonderfully. It does things like detect faces and brighten them up because that’s what people want to see in their photos”, he said.
As a concept, the idea of flagging AI-generated images might sound like an obvious step to reduce the spread of fake news catapulted by deepfake images and videos. However, identifying AI content and the intentions behind instances of image manipulation likely won’t be so cut and dry – especially for another robot like Gemini.
