American Sunlight Project Exposes AI-Driven Disinformation Network Manipulating Environmental and Tech Debates

New report reveals how a sophisticated foreign network is exploiting AI to distort online conversation

 

(Washington, DC) – The American Sunlight Project today released a new report exposing EcoBoost, an AI-powered disinformation network designed to manipulate online discussions on environmental and technology issues. The report reveals how EcoBoost has been operating across Western democracies — targeting primarily left-leaning X accounts with left-wing talking points — using artificial intelligence to generate fake social media personas that engage in deceptive online conversations and distort public perception.

 

ASP researchers identified 636 accounts within the EcoBoost network that have been active since at least mid-2024, producing AI-generated content in multiple languages to create the illusion of grassroots activism. These accounts flood social media with misleading narratives, particularly in response to discussions involving international governing bodies, NGOs, and academic institutions. While appearing to support pro-environmental causes, the network simultaneously fosters division across the political spectrum. The report also links EcoBoost to known Russian influence operations, including the Matryoshka network, with evidence suggesting it is operated by a malign Russian-language entity.

 

“The EcoBoost network is a stark reminder that AI-driven disinformation is no longer a future threat, it’s here now actively shaping public discourse,” said Nina Jankowicz, Co-founder and CEO of the American Sunlight Project. “Bad actors are using AI to manufacture credibility, hijack critical conversations, and manipulate public opinion at an unprecedented scale. Our findings show that disinformation knows no partisan bounds. Without urgent action from governments and tech platforms, these influence operations will continue to thrive in the shadows.”

 

The report details how EcoBoost uses AI-generated profile images and text to fabricate engagement, deceiving users into believing its narratives are widely supported. The network has engaged in mass-reply campaigns targeting discussions on climate change, AI policy, and environmental regulations, often amplifying misleading or divisive rhetoric. Some EcoBoost accounts have inadvertently posted Russian-language characters and code snippets, indicating ties to a foreign adversarial entity. The findings highlight major gaps in social media platforms’ ability to detect and counter AI-driven disinformation, a problem worsened by the rollback of Trust and Safety teams.

 

ASP’s findings raise serious concerns about the growing role of AI in disinformation campaigns and the inability of major platforms to keep pace with evolving threats. The report calls on policymakers, researchers, and tech companies to take immediate steps to improve transparency, restore researcher access to platform data, and implement stronger AI content moderation policies.

Next
Next

NEW REPORT: Russian propaganda may be flooding AI models