Spyke

Syndicated from the fediverse. Read and engage on the original instance.

View original on scribe.disroot.org

How AI-Driven Influence Operations Attack Ukraine

cross-posted from: https://scribe.disroot.org/post/10916724

When US military aid to Ukraine fell by 99% and Europe stepped into the breach in 2025, Russia’s influence machinery pivoted. Storm-1516, a network linked to the Internet Research Agency, had spent years spreading fabrications aimed at undermining international support for Ukraine. Now it set its sights on the Europeans picking up the tab.

Germany, suddenly Ukraine’s largest military donor at €9bn, and France, whose president had emerged as one of Kyiv’s leading champions, became the primary targets. According to a February 2026 NewsGuard investigation, Storm-1516 has published 34 false claims aimed at the two countries since January 2025, generating 274m views. The core message: Ukraine is bankrupting Europe, exhausting its citizens, and dragging the continent toward war.

...

Russia no longer operates alone, either. Moscow and Beijing have deepened coordination on artificial intelligence governance, digital standards, and media policy. This partnership now extends to influence operations ... Russian-linked actors using the Chinese AI model DeepSeek to conduct influence campaigns, while independent analysis showed Chinese AI models embed content controls that reproduce Kremlin narratives on Ukraine.

The European Commission’s 2025 report on foreign information manipulation concluded that Russian and Chinese actors have “fully implemented AI tools” for such operations.

...

As the London-based Institute for Strategic Dialogue and CIDOB, the Barcelona Centre for International Affairs, have documented, anti-enlargement narratives from Russian and Chinese-affiliated outlets circulate freely in Serbian, Bosnian, and Montenegrin media.

...

During Hungary’s April 2026 elections ...the ruling Fidesz party [also] ran a campaign built on Ukraine fatigue, reinforced by AI-generated imagery, deploying synthetic news anchors and deepfake endorsements on TikTok and content on X and Telegram.

But the propaganda failed. Hungarian voters saw through the lies and rejected Viktor Orbán in a historic landslide: Péter Magyar’s Tisza Party won 141 out of 199 seats, ending 16 years of Fidesz rule.

...

Hungary’s lesson is important: The most effective counter to AI-powered propaganda is not better regulation — it is better democratic speech. Credible counternarratives from independent media, fact-checkers, and political leaders willing to make the case for Ukraine in plain language did more to defeat Fidesz’s narrative machine than any platform rule. Scaling that kind of counter speech — investing in independent journalism, civic media literacy, and ensuring truthful voices are louder and faster than fabricated ones – is the front line.

...

Regulation still has a role, but a more modest one than Europe’s policymakers like to admit. Bringing Telegram under meaningful DSA [Digital Services Act] obligations, enforcing the AI Act’s transparency rules, and helping Western Balkan candidate countries build media literacy and fact-checking capacity can narrow the gaps that adversaries exploit. But the DSA’s record — two fines in over two years — is a reminder that rules on paper mean little without the will to apply them. What defeated the propaganda in Hungary was not an algorithm or a regulation. It was citizens who refused to believe the lies.

...

Archived

How AI-Driven Influence Operations Attack Ukrainehttps://cepa.org/article/how-ai-driven-influence-operations-attack-ukraine/Open linkView original on scribe.disroot.org
16

No replies yet

No comments on the original post yet.