Philippines, China clashes trigger money-making disinformation

MANILA, Philippines — Violent confrontations between Chinese and Philippine vessels in the South China Sea are being manipulated online by disinformation networks for profit, posing risks to regional security, an AFP investigation has found.

Since Philippine President Ferdinand Marcos took office in 2022 and began standing up to China over its territorial claims, social media has been awash with posts that often exaggerate the country's defense capabilities or allude to an imminent war in the region.

AFP uncovered a coordinated network of dozens of Facebook and YouTube channels that direct users to a bogus news website that appears to use artificial intelligence (AI) to rapidly churn out unfounded claims for advertising revenue.

Analysts say it is unclear if the network is directly linked to state actors, but the content often mirrors China's position on the dispute.

Together, the network's associated pages and channels have a combined following of more than 10 million people.

Elise Thomas, a senior analyst at London's Institute for Strategic Dialogue think tank, said revenue-making disinformation campaigns have emerged as a "big industry" in Southeast Asia, where labour is relatively cheap.

"It is a horrific business; a lot of it is built around preying on people," she said.

Some of the social media pages and websites claim to be American, but are in fact managed in Asian countries including Indonesia, Thailand, Bangladesh, India and Pakistan.

AFP found at least 25 Facebook "military interest" pages that misrepresented old photos and videos of army operations to falsely claim that Washington was helping its ally Manila prepare for war.

The posts take users to a bogus news site with a history of publishing fabricated articles and YouTube videos on the South China Sea dispute and the war in Ukraine.

In one, a digitally altered picture of Ukraine soldiers transporting crates falsely suggested that the US was sending Javelin anti-tank missiles to the archipelago.

Some were accidentally published with the headline: "I am a language model AI and cannot perform tasks that require real-time information".

One post debunked by AFP has been shared more than 25,000 times.