I firmly believe that artificial intelligence will not exacerbate the spread of misinformation.

Will AI Really Accelerate Disinformation? A Thoughtful Perspective

In recent discussions, a common concern has emerged: Will artificial intelligence amplify the spread of false information, leading to an unprecedented surge in disinformation campaigns? Many worry that as AI capabilities grow, so too will the volume of synthesized, misleading content flooding our digital spaces.

A key argument hinges on the amount of “AI-generated noise” already present on social media platforms. If we consider the content we encounter daily—be it TikTok videos, YouTube shorts, or other quick-format media—the prevalence of AI-produced material seems substantial, suggesting an inevitable rise in disinformation.

However, I maintain a different viewpoint. Suppose you give yourself or me a smartphone and challenge us to scroll through our preferred social media feeds. Without trying very hard, most of us would probably observe that we encounter approximately 100 to 150 videos or posts in a typical session. Whether the content is human-generated or AI-created, the quantity remains roughly constant. Introducing AI-generated content doesn’t inherently increase the volume of material we consume; it simply populates it differently.

While some argue that a larger influx of content inevitably means more misinformation, the reality is that humans already generate an overwhelming amount of disinformation at an astonishing scale. This existing flood has practically made it impossible for anyone to fully process or verify all the information they come across. In other words, adding more AI-generated disinformation doesn’t significantly alter the landscape because our exposure is already saturated.

Our media consumption habits tend to gravitate toward familiar formats—cat videos, humorous mishaps, emotional political commentary, and miscellaneous content curated by algorithms. Whether the videos are AI-generated or not, the proportion of disinformation in what I see remains relatively unchanged over recent years. My brain remains attuned to certain formats that lend themselves to political misinformation, and I doubt AI advances will drastically shift this dynamic.

It’s worth noting that AI-generated deception can sometimes be more subtle and insidious than blatant falsehoods. For instance, a manipulated clip of a politician, enhanced with clever editing and contextual framing, can be more persuasive and less obviously false than straightforward lies. An example might be a doctored audio or video clip circulating with a provocative quote, which can stir confusion or misinformation without blatantly flashing “fake” labels.

The most probable concern here is the proliferation of deepfake videos featuring celebrities or public figures saying things they never uttered. While this is a genuine challenge, I believe that against the vast torrent