Cuban warns of weekly deepfake scams as Meta logs $16bn in prohibited ad revenue
Internal documents show Meta expects billions from prohibited ads, while Mark Cuban warns of weekly AI deepfake victimisation targeting the elderly.
Key points
- Meta internal documents project approximately $16 billion in revenue from scam and prohibited product advertisements.
- Mark Cuban reports receiving weekly emails from victims of AI-generated deepfake ads using his likeness.
- Meta's automated systems require a 95% probability threshold before banning suspected scam advertisers.
Reuters reported on 6 November 2025 that internal Meta documents indicate the company expects roughly $16 billion in revenue from advertisements promoting scams and prohibited products.reported This figure represents about 10 per cent of Meta's total revenue for 2024.reportedInternal documents show Meta expects billions in revenue from ads promoting scams and prohibited products.Reuters reported that Meta's internal metrics project $16 billion in revenue from high-risk advertisements. The documents reveal that users across Meta's applications were seeing approximately 15 billion high-risk ads daily.reported These ads contained obvious markers of fraud, yet they remained active on the platform.
Evidence: reported
Cuban: Facebook knowingly runs AI deepfake scam ads; docs say $16B fraud revenueaifleet-hn-ai
Evidence: reported
Cuban: Facebook knowingly runs AI deepfake scam ads; docs say $16B fraud revenueaifleet-hn-ai
Evidence: reported
Cuban: Facebook knowingly runs AI deepfake scam ads; docs say $16B fraud revenueaifleet-hn-ai
How we got here
Mark Cuban, a billionaire investor, stated on an August episode of the Pivot podcast that he receives weekly emails from elderly individuals who have purchased a non-existent diabetes cure for $294.reported Cuban explained that these advertisements use AI-generated deepfakes of his face and voice. He noted that Meta continues to run these ads despite knowing they target vulnerable people. Some ads included disclaimers stating the content was AI-generated and for entertainment purposes.
Evidence: reported
Cuban: Facebook knowingly runs AI deepfake scam ads; docs say $16B fraud revenueaifleet-hn-ai
What is new
The Reuters report highlights specific internal metrics regarding Meta's tolerance for fraudulent advertising. A late 2024 document valued the high-risk ad category at approximately $7 billion annually. Furthermore, Meta's safety staff estimated in a May 2025 presentation that the company's platforms contribute to one-third of successful scams in the United States. These figures provide a quantified view of the scale of the issue.
What to watch
Developers should monitor how Meta's automated systems handle advertisers below the 95% scam probability threshold.reported According to the documents, suspected scammers below this threshold simply pay higher rates to keep their ads live.reported This mechanism allows fraudulent accounts to remain active while generating revenue for the platform. The company has not confirmed any changes to this policy.unverified
Evidence: reported
Cuban: Facebook knowingly runs AI deepfake scam ads; docs say $16B fraud revenueaifleet-hn-ai
Evidence: reported
Cuban: Facebook knowingly runs AI deepfake scam ads; docs say $16B fraud revenueaifleet-hn-ai
Availability and limits
AI application developers must consider the implications of Meta's 95% probability threshold for ad approval. This limit means that many fraudulent campaigns operate openly, potentially affecting brand safety and user trust. Meta has not confirmed any changes to this policy.unverified The documents state that suspected scammers below the 95% threshold pay more per ad and stay live.
Why it matters
- Meta's 95% probability threshold allows significant fraudulent activity to persist, impacting platform integrity.
- The $16 billion revenue projection indicates substantial financial incentives for maintaining current ad policies.
- Deepfake ads using celebrity likenesses demonstrate the ease of generating convincing synthetic media.
- Safety staff estimates suggest Meta's platforms facilitate a third of successful US scams.
Topics: Meta, deepfakes, advertising fraud, AI safety, Mark Cuban
Updated
Then / now / next
Meta's safety staff estimated platforms played a part in a third of US scams in May 2025.
Reuters reported on 6 November 2025 that internal documents show $16 billion expected from scam ads.
Developers should watch for any policy changes regarding the 95% probability threshold for ad bans.
Story timeline
- Then Meta's safety staff estimated platforms played a part in a third of US scams in May 2025.
- 05 Nov 2026 Cuban: Facebook knowingly runs AI deepfake scam ads; docs say $16B fraud revenue
- Now Reuters reported on 6 November 2025 that internal documents show $16 billion expected from scam ads.
- Next Developers should watch for any policy changes regarding the 95% probability threshold for ad bans.
The strongest doubt
Counterpoint from the sources
Meta has not publicly confirmed the details of the internal documents or the specific mechanics of its ad approval thresholds.
For readers
Developers
- Meta's 95% probability threshold allows many fraudulent ads to remain live.
- Deepfake technology is being used to create convincing scam advertisements.
Consumers
- Elderly users are targeted by AI-generated ads using celebrity likenesses.
- Some ads include disclaimers but still result in fraudulent purchases.
Investors
- Internal documents project $16 billion in revenue from scam-related ads.
- This figure represents approximately 10 per cent of Meta's 2024 revenue.
Ask the article
Local keyword search of this article; no generated answers.Did this article give you the evidence you needed?
Your choice stays on this device. No public tally.Questions readers ask
How much revenue does Meta expect from scam ads?
Internal documents project approximately $16 billion, or 10 per cent of 2024 revenue.
What threshold does Meta use to ban scam advertisers?
Automated systems must peg an account as at least 95% likely to be running a scam.
Does Meta know about the deepfake ads?
Mark Cuban claims Meta knows the ads are fraudulent but continues to run them.
Connected stories
Reading receipt
- Meta internal documents project approximately $16 billion in revenue from scam and prohibited product advertisements.
- Mark Cuban reports receiving weekly emails from victims of AI-generated deepfake ads using his likeness.
- Meta's automated systems require a 95% probability threshold before banning suspected scam advertisers.
Still open: Will Meta change its 95% probability threshold for banning scam advertisers?
Still open: How many users have been affected by the deepfake ads using Mark Cuban's likeness?