Artificial intelligence dominates social media, but only 44% of users are confident they can spot it, CNET finds

Amnesty International slope It has affected all social media platforms, from Soulless pictures to Strange videos and Surface text. The vast majority of US adults who use social media (94%) believe they encounter content generated or edited by AI, but only 44% of US adults say they are confident in their ability to tell real photos and videos from AI-generated photos and videos, according to an exclusive CNET poll. This is a big problem.

There are a lot of different ways people resist AI content. Some solutions are focusing on better labels for AI-generated content, as it’s harder than ever to trust our eyes. Of the 2,443 respondents who use social media, half (51%) believe we need better online AI taxonomies. Others (21%) believe there should be a complete ban on AI-generated content on social media. Only a small group (11%) of respondents say they find AI content useful, informational, or entertaining.

AI is not going anywhere, and it is fundamentally reshaping the Internet and our relationship with it. Our survey shows that we still have a long way to go to reckon with it.

Key findings

• Most US adults who use social media (94%) believe they encounter AI content on social media, but far fewer (44%) can confidently distinguish between real and fake photos and videos.
• Many US adults (72%) said they take action to determine if a photo or video is real, but some don’t do anything, especially among Boomers (36%) and Generation X (29%).
• Half of US adults (51%) believe that AI-generated and edited content needs to rank better.
• One in five (21%) believe AI content should be banned on social media, with no exceptions.

Adults in the United States don’t feel they can explore AI media

Vision no longer believes in the age of artificial intelligence. Tools like OpenAI Sora video generator and google Nano banana An image model can create highly realistic media, with chatbots seamlessly compiling blocks of text that look like a real person wrote them.

So it’s understandable that a quarter (25%) of US adults say they’re not confident in their ability to differentiate between real photos and videos and those produced by artificial intelligence. Older generations, including Boomers (40%) and Generation X (28%), are the least trusting. If people don’t have a lot of knowledge or exposure to AI, they will likely feel uncertain about their ability to accurately detect AI.

People take action to verify content in different ways

The ability of artificial intelligence to simulate real life makes it even more important to verify what we see online. Nearly three in four American adults (72%) said they take some form of action to determine if a photo or video is real when they raise their suspicions, with Generation Z being the most likely (84%) of the age groups to do so. The most obvious — and most common — method is to closely examine photos and videos for visual cues or artefacts. More than half of US adults (60%) do so.

But AI innovation is a double-edged sword; Models have improved rapidly, eliminating the previous errors we used to rely on to detect AI-generated content. The em dash has never been a reliable sign of AI, but extra fingers in photos and continuity errors in videos were once prominent red flags. Newer AI models typically don’t make these mistakes. So we all have to work harder to determine what is real and what is fake.

As visual indicators of AI disappear, other forms of content verification are increasingly important. The next two most common methods are checking ratings or disclosures (30%) and searching for content elsewhere online (25%), such as on news sites or through a reverse image search. Only 5% of participants reported using a com. deepfake Detection tool or location.

But 25% of US adults don’t do anything to determine whether the content they see online is real. This lack of work is highest among Boomers (36%) and those in Generation X (29%). This is worrying, as we have already seen that AI is an effective tool for achieving this Abuse and scam. Understanding the origins of a post or content is an important first step to navigating the Internet, where anything can be faked.

Half of US adults want better nomenclature for artificial intelligence

Many people are working on finding solutions to deal with the onslaught of artificial intelligence. Labeling is a major area of ​​opportunity. The classification relies on social media users to disclose that their posts were created with the help of artificial intelligence. This can also be done behind the scenes via social media platforms, but it is somewhat difficult, leading to random results. That’s likely why 51% of US adults believe we need better classification of AI content, including deepfakes. Support was strongest among Millennials and Generation Z, at 56% and 55%, respectively.

Other solutions aim to control the flow of AI content being shared on social media. All major platforms allow AI-generated content, as long as it doesn’t violate their general content guidelines – for example, nothing illegal or offensive. But some platforms have introduced tools to limit the amount of AI-generated content you see in your feeds; Pinterest rolled Its filters Last year, while TikTok was still around a test Some of their own. The idea is to give everyone the ability to allow or exclude AI-generated content from their feeds.

But 21% of respondents believe that AI content on social media should be banned completely, with no exceptions allowed. This number is highest among Generation Z at 25%. When asked if they thought AI content should be allowed but strictly regulated, 36% said yes. These low percentages can be explained by the fact that only 11% find AI content provides meaningful value – whether entertaining, informational or useful – and 28% say it provides little or no value.

How to limit AI content and detect potential deepfakes

Your best defense against AI deception is to be keen-eyed and trust your intuition. If something is too weird, too brilliant, or too good to be true, it probably is. But there are other steps you can take, such as using a deepfake detection tool. There are many options. I recommend starting with Content Authenticity Initiativetool, because it works with many different file types.

You can also check the account that shared the post for red flags. Often times, AI-based ramps are shared by bulk ramp producers, and you’ll easily be able to see this in their feeds. It will be full of strange videos that don’t seem to have any continuity or similarity between them. You can also check to see if anyone you know is following them or if this account is not following anyone else (this is a red flag). Spam posts or fraudulent links are also indicators that an account is not legitimate.

If you want to limit the AI ​​content you see in your social feeds, check out our guides to turning it off or muting Meta AI in Instagram and Facebook And liquidation Artificial Intelligence posts on Pinterest. If you encounter an error, you can mark the post as something you don’t care about, which should signal to the algorithm that you don’t want to see more likes of it. Outside of social media, you can disable it Apple intelligenceArtificial intelligence in pixel and Galaxy Phones and gemini in Search Google, Gmail, and Docs.

Even if you do all this and still get scammed by the AI, don’t feel bad about it. There is little we can do as individuals to fight the tide of AI decline. We’re all likely to make mistakes sometimes. Until we have a global system for effective AI detection, we have to rely on the tools we have and our ability to educate each other about what we can do now.

methodology

CNET commissioned the poll from YouGov Plc. All figures, unless otherwise stated, are from YouGov Plc. The total sample size was 2,530 adults, of whom 2,443 used social media. Fieldwork was conducted from February 3 to 5, 2026. The survey was conducted online. Figures have been weighted and represent all US adults (age 18+).