- Deepfake Watch
- Posts
- Deepfake Watch 8
Deepfake Watch 8
Facebook Has A “WTF AI” Problem | Deepfake Watch
Browser View | March 22, 2024 | Subscribe
This week we shall look at the following cases:
What on earth is this Shrimp Jesus epidemic on Facebook?
Chinese social media users are resurrecting dead artists; the parents of one revived singer are not happy about it
Giorgia Meloni is seeking €100k in damages for deepfake porn defamation
A US tech policy non-profit simulated deepfake disruption scenarios for the upcoming polls, highlighting several challenges the country is expected to face
Deepfake alert: video of AIMIM leader Asaduddin Owaisi chanting Hindu mantra
Opt-in to receive the newsletter every Friday.
Infestation of the bizarre
Paintings of Julius Caesar with fake Caesar quotes, images of MCU and DC superheroes in Ancient Egypt, videos showing Harry Potter characters as Sicilian mafiosos.
For a year I’ve been seeing more and more such AI-generated content flock to my Instagram and Facebook feeds. Some of these photos and videos are likely to make you go “WOW”. But now, it’s starting to get a little “WTF”.
Jesus as a composite of seahorses? WTF!
Researchers at Stanford Internet Observatory found scammers and spammers exploiting the eyeball-grabbing potential of AI to draw unsuspecting users and sending them to shady websites.
They studied 120 pages, some of which were run by the same administrators, and detected several scam pages that directed users to websites selling non-existent items, or requesting personal data. Some others redirected to low-ranking websites riddled with Google ads.
The pages leveraged AI to create images that were likely to be rated high on engagement metrics by Facebook’s algorithms, and eventually get pushed into the feeds of unwitting users.
Looking through the comments of such posts, they found that users were either unaware, or indifferent to the synthetic origins of the photos. Many users were found giving an “Amen” to these crustacean Jesuses.
Some of these pages also deployed other deceptive tactics, such as hijacking the Facebook page of a highschool band.
Meta’s AI-labeling policies are ineffective - they can only detect and label content that contain ‘industry-standard indicators’ of being AI. As for the rest - it seems to be open season.
AI Resurrection Fever Strikes China
Chinese singer and actor Qiao Renliang died in 2016 of suicide. But he recently appeared in a video on Chinese social media, saying, “Actually, I never really left.”
Renliang was one of the many, along with Chinese-American singer Coco Lee, and Taiwanese-Canadian actor Godfrey Gao, to be resurrected using generative AI.
Their videos have started going viral on Douyin (the Chinese name for TikTok) and have drawn the disapproval of the family members of the resurrected
"They did not ask for our consent. It was my niece who saw the video and sent it to me. This is just exposing old scars," Renliang’s father told Hong-Kong-based newspaper Oriental Daily.
While speaking to Chengdu-based Hongxing News, he added, “Who gave them the right to do this? It’s not just about (legal) infringements; it’s a lack of respect. We cannot accept this if it’s for commercial gain.”
Giorgia Meloni seeks €100,000 for deepfake porn defamation
Italian PM Giorgia Meloni is seeking €100,000 (₹90 lakh) in damages for deepfake porn videos of her being circulated online.
A 40-year-old man and his 73-year-old father are currently being investigated for producing the videos in question, and are likely to face criminal defamation charges. Investigators stated that they tracked down the suspects through the mobile phone used to publish the videos.
The Italian PM’s legal team stated that if the request for damages is successful, she will donate the amount to a fund to support female victims of male violence.
War-gaming deepfake disruptions to 2024 US elections
In New York, several former senior US officials, civil society leaders and tech executives were gathered to carry out a simulated exercise called the “The Deepfake Dilemma”, to rehearse for the upcoming US elections later this year.
Carried out by Washington-based non-profit The Future US, the exercise highlighted how generative AI tools are set to boost election-related disinformation in a country that is already grappling with the chaos of the ‘stolen election’ conspiracy theories that led to the January 6 Capitol Hill riots in 2021.
The following scenarios were some of the many that were simulated during the exercise:
“It’s Election Day in Arizona and elderly voters in Maricopa County are told by phone that local polling places are closed due to threats from militia groups”.
“Meanwhile, in Miami, a flurry of photos and videos on social media show poll workers dumping ballots”
The exercise highlighted the challenge posed by US’ decentralised election commissions, without any indication of who’s in charge at a country-wide level.
Nick Penniman, CEO of bipartisan organisation Issue One, told NBC News that during the exercise, “because patterns of working together haven’t formed, few people understood exactly how they should be coordinating with others or not.”
Asaduddin Owaisi Deepfake Alert
There is a video being shared on social media showing AIMIM leader Asaduddin Owaisi chanting a Hindu mantra fluently.
The video appeared sketchy to us, so we ran it by some media forensics experts. They confirmed to us that the video is indeed a deepfake.
AI was used to sync Owaisi’s lip movement to the overlaid audio. The grainy nature of the lip movement highlighted a poor quality deepfake.
Have you been a victim of AI?
Have you been scammed by AI-generated videos or audio clips? Did you spot AI-generated nudes of yourself on the internet?
Decode is trying to document cases of abuse of AI, and would like to hear from you. If you are willing to share your experience, do reach out to us at [email protected]. Your privacy is important to us, and we shall preserve your anonymity.
About Decode and Deepfake Watch
Deepfake Watch is an initiative by Decode, dedicated to keeping you abreast of the latest developments in AI and its potential for misuse. Our goal is to foster an informed community capable of challenging digital deceptions and advocating for a transparent digital environment.
↗️ Was this forwarded to you? Subscribe Now
We invite you to join the conversation, share your experiences, and contribute to the collective effort to maintain the integrity of our digital landscape. Together, we can build a future where technology amplifies truth, not obscures it.
For inquiries, feedback, or contributions, reach out to us at [email protected].
🖤 Liked what you read? Give us a shoutout! 📢
↪️ Become A BOOM Member. Support Us!↪️ Stop.Verify.Share - Use Our Tipline: 7700906588
↪️ Follow Our WhatsApp Channel↪️ Join Our Community of TruthSeekers
Copyright (C) " target="_blank">unsubscribe