Synthetic Media and Bots Outpace Human Verification

The Rise of Synthetic Media and Bot Traffic

The proliferation of synthetic media has surged since 2025, with AI-generated content now surpassing human verification capabilities. An Iran-linked outlet, Explosive News, has exemplified this trend by producing two-minute Lego-style propaganda videos within 24 hours. These synthetic segments, designed to mimic real events, are intended to spread rapidly before fact-checkers can respond. The White House has similarly adopted such tactics, releasing vague ‘launching soon’ videos that were later revealed to promote its official app. This shift reflects how institutional communication has embraced the aesthetics of leaks and virality, blurring the distinction between official statements and synthetic narratives.

Automated Traffic and the Virality Paradox

“Open source is not the place to determine what did or did not happen.”

Automated traffic now constitutes 51% of internet activity—up from 49% in 2024—with bad bots accounting for 37% of this volume. These bots prioritize engagement over accuracy, flooding platforms with low-quality content that outpaces human verification. According to the 2026 State of AI Traffic & Cyberthreat Benchmark Report, automated systems dominate internet traffic, favoring virality over factual accuracy. Henk van Ess, an investigative trainer, noted, ‘Generative media breaks the assumption that an image is a record of something.’ This has created a digital environment where authenticity is no longer guaranteed, and the speed of dissemination exceeds the speed of verification.

Erosion of Traditional Verification Methods

A significant shift in digital verification has occurred: the absence of a digital footprint no longer signals authenticity. Previously, a lack of metadata or camera trails implied originality, but now it may indicate the content was never captured. This inversion has led to a paradox where truth lags behind engagement. Automated systems, which dominate traffic, prioritize virality over accuracy, ensuring synthetic records spread before they can be contested. Maryam Ishani, an OSINT journalist, explained, ‘We’re perpetually catching up to someone pressing repost without a second thought.’ The algorithm’s preference for rapid sharing means open-source investigators are consistently behind, struggling to keep pace with synthetic content.

“Generative media breaks the assumption that an image is a record of something.”

Challenges in Detecting Synthetic Media

Traditional verification methods are also being eroded. For instance, Planet Labs, a commercial satellite provider, has restricted access to imagery of Iran and the Middle East conflict zone at the request of the U.S. government. Defense Secretary Pete Hegseth dismissed open-source verification as a tool for determining factual events, stating, ‘Open source is not the place to determine what did or did not happen.’ This restriction limits the ability to independently verify events, creating a vacuum where generative AI defines what is perceived as real.

Sector-Specific Bot Activity

Bot traffic varies by sector, with retail experiencing the highest levels at 59%, followed by gaming, where bad bots account for over 57% of traffic. In contrast, sectors like entertainment and government report lower bot activity, though advanced bad bots still dominate up to 78% of bot activity. These variations underscore the need for sector-specific strategies to combat synthetic media and automated content manipulation. For example, in retail, bots are often used for price scraping and fake reviews, while in gaming, they drive up demand for in-game items through automated purchases.

Strategies for User Verification

Experts like Henk van Ess recommend a multi-step verification process. Users should look for cinematic elements—images that are overly composed, evenly lit, or staged, which are rare in real-world scenarios. Running multiple reverse-image searches using tools like Google Lens, Yandex, and TinEye can uncover different histories of an image, even if it lacks a digital footprint. Inspecting peripheral details such as parking signs, manhole covers, or shadow angles can reveal inconsistencies that synthetic content often overlooks. Treating detection tools as prompts rather than verdicts is crucial, as confidence scores without explanations are insufficient for determining authenticity. The free tool ImageWhisperer combines these signals to help users verify content more effectively.