Meta and OpenAI announced that they had disrupted covert online influence operations run by an Israeli company (STOIC ) , amid the war in Gaza.
The tech giants announced that STOIC, a Tel Aviv-based political marketing and business intelligence firm, is developing its products and tools to manipulate political conversations online.
See also: War in Gaza: Phishing campaign deceives victims with alleged donations
OpenAI said Thursday that it had banned a network of accounts affiliated with STOIC, accusing them of posting anti-Hamas and pro-Israel content and acting as an “Israeli recruitment threat agent.”.
The accounts leveraged OpenAI models to spread misinformation about the war in Gaza as well as the upcoming elections in India.
Specifically, they leveraged artificial intelligence models to create articles and comments, which were published on multiple platforms, mainly on Instagram , Facebook , X and websites associated with the feature, according to OpenAI.
"Some of the campaigns we stopped exploited our models to create a false image on social media. For example, they created replies to their own posts in order to present a false image of online interest," he said.
However, according to OpenAI, the network's activity "appears to have attracted little to no interest, beyond its own inauthentic accounts.".
See also: Israel-Hamas: How Social Media Amplifies Disinformation
STOIC is described as an AI content creation system that helps users automatically produce targeted content and distribute it quickly and organically across relevant platforms.
Meta confirmed in a quarterly security on Wednesday that it removed over 500 Facebook accounts, one group and 11 pages, as well as more than 30 Instagram accounts, which were linked to the same influencer business.
It reported that fake accounts posing as Jewish students, African Americans and other concerned citizens were targeting audiences in the US and Canada as part of a covert campaign linked to STOIC.
“There are many examples on these networks of how they are using artificial intelligence tools to produce content. These tools may allow them to create content faster or with greater intensity. However, they have not affected our ability to detect them,” Meta’s head of threat , Mike Dvilyansky, told Reuters.
Read more: Social media denies Shadowban regarding Palestine
Facebook's parent company announced that it has banned STOIC and sent a letter demanding the immediate termination of activities that violate Meta's policies.
Source: independent
