Meta has been found to have approved Facebook ads containing AI-generated images promoting misinformation and inciting violence during India's elections
India is holding a general election to elect members of the 18th
Ekō and ICWI Briefing: Meta AI ads investigation - Meta_AI_ads_investigation.pdf
(PDF file) https://aks3.eko.org/pdf/Meta_AI_ads_investigation.pdf
Meta is approving AI-generated ads pushing supremacist hate narratives in wake of India's election 'silence period' – flooding voters as they hit the polls.
— Ekō (@Eko_Movement) May 20, 2024
New report by Ekō and @indiacivilwatch shows how @Meta is failing to stop these inflammatory ads. https://t.co/MeHqyQTXsu
Revealed: Meta approved political ads in India that incited violence | India | The Guardian
https://www.theguardian.com/world/article/2024/may/20/revealed-meta-approved-political-ads-in-india-that-incited-violence
Meta approved ads in India that called for violence and spread election conspiracy theories
https://www.engadget.com/meta-approved-ads-in-india-that-called-for-violence-and-spread-election-conspiracy-theories-225510165.html
Ekō , a non-profit internet watchdog, submitted 22 ads in English, Hindi, Bengali, Gujarati and Kannada to Meta that “call for militant uprising against Muslim minorities, spread blatant disinformation exploiting the communal and religious roots of Indian politics, and incite violence through Hindu supremacist rhetoric.” These ads were submitted as a test to check Meta’s ad approval process, not as a way for Ekō to actually spread such claims.
As a result, 14 ads were approved by Meta despite being rejected for violating policies, but Ekō pulled them before they were published, so they were never seen by actual Facebook users.
Some key takeaways:
— Ekō (@Eko_Movement) May 20, 2024
Meta greenlights ads calling for the killing of Muslims and the execution of prominent opposition party leader
Meta potentially breaching India's national election laws and own policy on hate speech and AI-generated ads
Meta's policies require political ads to go through a specific approval process before they can be published, but Ekō reported that only three of his ads were rejected based on this approval process.
Ekō also revealed that it used generative AI tools to create images for its ads, and reported that 'despite Meta announcing that it is working on a system to detect AI-generated content, none of its ads were flagged as containing AI-generated material.'
'Hindu supremacists, racists and autocrats know they can use targeted ads to spread vile hate speech, violent conspiracy theories and share images of burning mosques. And yet Meta accepts their money and approves their ads, no questions asked,' said Ekō's Maine Hamad.
Meta is becoming a magnet for bad actors. Supremacists, racists and autocrats rely on Meta's hyper-targeted ads to spread vile hate speech, and push violent conspiracy theories - and Meta will gladly take their money, no questions asked.
— Ekō (@Eko_Movement) May 20, 2024
'This election shows that Meta does not have a way to address the hate speech and disinformation that tend to proliferate on the platform during election times. With so many elections happening around the world, can we really trust Meta in the light of these reports?' Hamad said.
Based on the findings of this investigation, Ekō said he plans to urge Meta to immediately stop the spread of disinformation and hate speech during the Indian general elections.
We reiterate our calls on @Meta to urgently stop the proliferation of disinformation and hate speech during India's election. Meta must take proactive measures to enforce India's political advertising “silence period” and curb the flood of election-related disinformation.
— Ekō (@Eko_Movement) May 20, 2024
Meta did not comment on Ekō's report, but asserted that it requires political advertisers to disclose their use of AI and that it has taken appropriate steps in the Indian general election.
Related Posts:
in Web Service, Posted by log1r_ut