Back in May, Google announced it was working on a feature called “about this image” that gives users verified data regarding any photo on the internet. Well, it just rolled out as part of search, so you won’t be able to get away with passing off somebody else’s photo of a 1988 Burger King Alf plushie as your own.
Here’s how it works. Just use Google Search, select an image and click on the three dots on the right-hand corner to access the tool. You’ll receive a whole gob of useful information, including when the image was originally published, if it’s been published since then and where it’s popped up throughout the years. A veritable cornucopia of metadata.
Google
The obvious use case scenario for this is verifying whether or not an image used to accompany a news event is legit, or if it’s been taken out of context from something that happened in 2007 to drum up misinformation. To that end, the tool also shows you how other sites use and describe the image, similar to how search already handles factual information via the “perspectives” filter and the “about this result” tab. Google says you can also access the feature by clicking on the “more about this page” link, with more options to come.
Of course, there’s a little thing sweeping the world right now called artificial intelligence. The images generated by AI platforms can be tough to distinguish from the genuine article, so Google’s tool also lets you know if an image has been AI-generated or not. However, this depends on the metadata including this information, so the original image creators would have to opt-in. Google says its own AI-generated images will always feature the appropriate metadata.
That’s not the only tool Google’s rolling out to provide increased nuance for image searches. Fact Check Explorer, a handy app for journalists, will soon expand to include images. As for non-image based searches, the tech giant also announced software that creates AI-generated descriptions of websites, helping users research lesser-known entities.
This article originally appeared on Engadget at https://www.engadget.com/googles-ne...and-sniff-out-ai-fakes-165339778.html?src=rss
Here’s how it works. Just use Google Search, select an image and click on the three dots on the right-hand corner to access the tool. You’ll receive a whole gob of useful information, including when the image was originally published, if it’s been published since then and where it’s popped up throughout the years. A veritable cornucopia of metadata.
The obvious use case scenario for this is verifying whether or not an image used to accompany a news event is legit, or if it’s been taken out of context from something that happened in 2007 to drum up misinformation. To that end, the tool also shows you how other sites use and describe the image, similar to how search already handles factual information via the “perspectives” filter and the “about this result” tab. Google says you can also access the feature by clicking on the “more about this page” link, with more options to come.
Of course, there’s a little thing sweeping the world right now called artificial intelligence. The images generated by AI platforms can be tough to distinguish from the genuine article, so Google’s tool also lets you know if an image has been AI-generated or not. However, this depends on the metadata including this information, so the original image creators would have to opt-in. Google says its own AI-generated images will always feature the appropriate metadata.
That’s not the only tool Google’s rolling out to provide increased nuance for image searches. Fact Check Explorer, a handy app for journalists, will soon expand to include images. As for non-image based searches, the tech giant also announced software that creates AI-generated descriptions of websites, helping users research lesser-known entities.
This article originally appeared on Engadget at https://www.engadget.com/googles-ne...and-sniff-out-ai-fakes-165339778.html?src=rss
