The world’s leading search engines are currently under fire for some shocking behavior that was made public by NBC News recently.
DuckDuckGo, Google, and Bing are being called out for adding AI Deepfake porn toward the top of their respective search engine results. This has really highlighted some major concerns regarding the use and easy access of AI-based porn.
The videos in question are explicit and produced via real porn material. The only change is that they alter the actor’s face to make the individual seem as real as possible like perhaps adding a celeb. This is done through automated means by AI and in the end, users making the product are just required to feed systems with AI alongside images of the victim.
As per tests carried out online by media outlet NBC News, a whopping 36 celebs of female origin made use of a combo of names and words such as deepfakes. Seeing the explicit results appear at the top of the search is beyond shocking, one person added.
Google ended up returning 34 of the most searched results and then Bing returned 35. Similar problems were seen with DuckDuckGo as it had issues that were very much like that. The problem’s scale was yet to be specified.
Meanwhile, NBC News complained how relevant articles regarding the matter of nonconsensual deepfake porn were displaying these illicit pictures before the actual search results demanded by the user. This was even more common when the term fake nudes was used.
It’s worth mentioning how NBC news first switched off safety features for Bing and Google to see the effect arise when no safeguard was in place. Both search engines are known for their safety protection levels which work on default to give the browser the best and safest search experience.
Meanwhile, Neowin also carried out its specific tests on this front and proved how explicit images weren’t displayed for Bing in the default version when deepfakes were searched. However, it was alarming to witness how it did link certain webpages to inappropriate content to produce the AI deepfakes.
Google displayed explicit images when default was activated but the pictures were thankfully blurred and only displayed completely after users were notified of the picture’s explicit nature. Meanwhile, users were also given the chance to press the button willingly if they wished to see the real image.
The ease with which users are getting access to porn driven by AI technology is worth a mention and it’s continuing to be a serious problem in today’s day and age. Anyone looking for such content would turn off any built-in features designed for protection.
Such behavior needs to be stopped by adding advanced tech safeguards that might include suppressing pages promoting this content from the search or simply barring them as a whole.
This would drastically limit the harmful aspect linked to user search as more than 50% of the top search results on tests comprised of links to famous deepfake websites.
Software giant Microsoft was the only firm that failed to respond to requests for comments by NBC News. The others are taking the matter seriously and do realize how big of an alarming sign it is to see such acts take place.
Such news just goes to show how much harm AI can provide and how the technology is being used for the sake of non-ethical and malicious means. For months, we’ve been seeing how Deepfakes continue to serve as a major issue in other domains as well. Common examples include the world of politics.
Tech giants like Microsoft are one such firm that hopes to come up with software to detect deepfakes with ease and to stop people from misusing technology before the election period commences.
On the other hand, the FTC is looking for the best ideas to find audio recordings that are fake or created through AI. There is a competition going on with a staggering $25k price for the winner who can find a solution for the Voice Cloning Challenge. And the deadline is today.
Photo: Digital Information World - AIgen/HumanEdited
Read next: FTC Makes History After Barring US Data Broker Firm Outlogic From Selling User Location Data
DuckDuckGo, Google, and Bing are being called out for adding AI Deepfake porn toward the top of their respective search engine results. This has really highlighted some major concerns regarding the use and easy access of AI-based porn.
The videos in question are explicit and produced via real porn material. The only change is that they alter the actor’s face to make the individual seem as real as possible like perhaps adding a celeb. This is done through automated means by AI and in the end, users making the product are just required to feed systems with AI alongside images of the victim.
As per tests carried out online by media outlet NBC News, a whopping 36 celebs of female origin made use of a combo of names and words such as deepfakes. Seeing the explicit results appear at the top of the search is beyond shocking, one person added.
Google ended up returning 34 of the most searched results and then Bing returned 35. Similar problems were seen with DuckDuckGo as it had issues that were very much like that. The problem’s scale was yet to be specified.
Meanwhile, NBC News complained how relevant articles regarding the matter of nonconsensual deepfake porn were displaying these illicit pictures before the actual search results demanded by the user. This was even more common when the term fake nudes was used.
It’s worth mentioning how NBC news first switched off safety features for Bing and Google to see the effect arise when no safeguard was in place. Both search engines are known for their safety protection levels which work on default to give the browser the best and safest search experience.
Meanwhile, Neowin also carried out its specific tests on this front and proved how explicit images weren’t displayed for Bing in the default version when deepfakes were searched. However, it was alarming to witness how it did link certain webpages to inappropriate content to produce the AI deepfakes.
Google displayed explicit images when default was activated but the pictures were thankfully blurred and only displayed completely after users were notified of the picture’s explicit nature. Meanwhile, users were also given the chance to press the button willingly if they wished to see the real image.
The ease with which users are getting access to porn driven by AI technology is worth a mention and it’s continuing to be a serious problem in today’s day and age. Anyone looking for such content would turn off any built-in features designed for protection.
Such behavior needs to be stopped by adding advanced tech safeguards that might include suppressing pages promoting this content from the search or simply barring them as a whole.
This would drastically limit the harmful aspect linked to user search as more than 50% of the top search results on tests comprised of links to famous deepfake websites.
Software giant Microsoft was the only firm that failed to respond to requests for comments by NBC News. The others are taking the matter seriously and do realize how big of an alarming sign it is to see such acts take place.
Such news just goes to show how much harm AI can provide and how the technology is being used for the sake of non-ethical and malicious means. For months, we’ve been seeing how Deepfakes continue to serve as a major issue in other domains as well. Common examples include the world of politics.
Tech giants like Microsoft are one such firm that hopes to come up with software to detect deepfakes with ease and to stop people from misusing technology before the election period commences.
On the other hand, the FTC is looking for the best ideas to find audio recordings that are fake or created through AI. There is a competition going on with a staggering $25k price for the winner who can find a solution for the Voice Cloning Challenge. And the deadline is today.
Read next: FTC Makes History After Barring US Data Broker Firm Outlogic From Selling User Location Data
