Twitter, Meta Take Down ‘US-Based’ Propaganda Campaign

Researchers have brought to light a social media propaganda campaign on Twitter, Instagram, Facebook and other platforms aimed at furthering the interests of the US and its allies on the international stage.

The campaign is the first major covert pro-US propaganda effort taken down by the social media companies, said social media analysis firm Graphika and the Stanford Internet Observatory (SIO).

Dozens of accounts used by the campaign were removed from social media platforms in July and August.

US interests

The campaign is unusual in that it makes use of techniques similar to those used by the US’ adversaries, such as false social media accounts with AI-generated profile images.

The researchers said Twitter had identified the US and the UK as the “presumptive countries of origin”, while Facebook and Instagram parent Meta said the US was the “country of origin”.

But the report stopped short of identifying those countries’ governments as being behind the campaign, with the SIO saying it didn’t have the necessary information to attribute the activity to a single country or organisation.

“We believe this activity represents the most extensive case of covert pro-Western (influence operations) on social media to be reviewed and analysed by open-source researchers to date,” wrote Graphika and the SIO in their report.

Foreign influence

The accounts in question made use of profile images generated by an AI technique called generative adversarial networks (GANs), which assembles a visual image based on patterns recognised from other images and tries to make it pass as authentic when analysed by a rival AI.

The resulting images may look superficially realistic to a human observer, but often include anomalies that allow them to be spotted as fake.

The propaganda campaign targeted individuals in Asia and the Middle East with fake accounts that posed as independent journalists and posted short videos, memes, hashtag campaigns and petitions.

The campaign gained little interest from its intended audience, with most posts picking up only a “handful” of likes or retweets and less than one-fifth of accounts having more than 1,000 followers.

‘Inauthentic tactics’

The researchers said the lack of impact showed the “limitations of relying on inauthentic tactics”.

Most of the accounts began operating in 2019, with a few launched on Twitter in late 2017, the report said.

Matthew Broersma

Matt Broersma is a long standing tech freelance, who has worked for Ziff-Davis, ZDnet and other leading publications

Recent Posts

Google, DOJ Closing Arguments Clash Over Search ‘Monopoly’

Google clashes with US Justice Department in closing arguments as government argues Google used illegal…

5 hours ago

Stanford AI Scientist Working On ‘Spatial Intelligence’ Start-Up

Prominent Stanford University AI scientist Fei-Fei Li reportedly completes funding round for start-up based on…

6 hours ago

Apple Shares Surge Ahead Of New AI Hardware Launches

Apple shares surge on optimism that new AI-focused hardware launches will drive renewed sales, starting…

6 hours ago

Biden Vetoes Republican Measure In Row Over Contractors’ Unions

Biden vetoes Republican-backed measure amidst dispute over 'joint employer' status for contract workers, affecting tech…

7 hours ago

Lawyers Say Strict Child Controls In China Show TikTok Could Do Better

Lawyers in US social media addiction action say strict controls on Douyin in China show…

7 hours ago

London Black Cabs Sue Uber In Latest Legal Tangle

More than 10,000 London black cab drivers sue Uber claiming company acted illegally to obtain…

8 hours ago