Description: 143 deepfake ads, over 100 of which reportedly impersonated former British Prime Minister Rishi Sunak, were promoted on Meta's platform to advertise the fraudulent investment scheme "Quantum AI." Funding for the ads reportedly originated from 23 countries. Up to 462,000 users may have been exposed to the false content. The campaign used generative AI tools to create high-quality misinformation, including spoofed BBC news clips for added legitimacy
Editor Notes: Reconstructing the timeline of events: (1) August 12, 2023: Reported start of the deepfake ad campaign on Meta; (2) August 1, 2024: End of the one-month investigation period during which Fenimore Harper identified 143 deepfake ads; (3) January 13, 2024: Fenimore Harper publishes its findings. Read the full report here: https://www.fenimoreharper.com/s/FENIMORE-HARPER-REPORT_-DEEP-FAKED-POLITICAL-ADS-V2.pdf. The full list of 143 identified ads by Fenimore Harper can be accessed here: https://www.fenimoreharper.com/s/Deepfake-Finance-Scam_-Full-List-of-Ads.xlsx.
Entities
View all entitiesAlleged: Unknown deepfake technology developers developed an AI system deployed by Quantum AI scammers, which harmed Rishi Sunak , Quantum AI victims , Meta users and BBC News presenters.
Alleged implicated AI system: Unknown deepfake technology apps
Incident Stats
Risk Subdomain
A further 23 subdomains create an accessible and understandable classification of hazards and harms associated with AI
4.3. Fraud, scams, and targeted manipulation
Risk Domain
The Domain Taxonomy of AI Risks classifies risks into seven AI risk domains: (1) Discrimination & toxicity, (2) Privacy & security, (3) Misinformation, (4) Malicious actors & misuse, (5) Human-computer interaction, (6) Socioeconomic & environmental harms, and (7) AI system safety, failures & limitations.
- Malicious Actors & Misuse
Entity
Which, if any, entity is presented as the main cause of the risk
Human
Timing
The stage in the AI lifecycle at which the risk is presented as occurring
Post-deployment
Intent
Whether the risk is presented as occurring as an expected or unexpected outcome from pursuing a goal
Intentional
Incident Reports
Reports Timeline
More than 100 deepfake video advertisements impersonating Rishi Sunak were paid to be promoted on Facebook in the last month alone, according to research that has raised alarm about the risk AI poses before the general election.
The adverts…
EXECUTIVE SUMMARY
-
Over 100 deep-fake video advertisements impersonating Prime Minister Rishi Sunak were paid to be promoted on Meta's platform in the last month.
-
These ads may have reached over 400,000 people, despite explicitly breakin…
Variants
A "variant" is an incident that shares the same causative factors, produces similar harms, and involves the same intelligent systems as a known AI incident. Rather than index variants as entirely separate incidents, we list variations of incidents under the first similar incident submitted to the database. Unlike other submission types to the incident database, variants are not required to have reporting in evidence external to the Incident Database. Learn more from the research paper.
Similar Incidents
Did our AI mess up? Flag the unrelated incidents
Deepfake Obama Introduction of Deepfakes
· 29 reports
Defamation via AutoComplete
· 28 reports
Similar Incidents
Did our AI mess up? Flag the unrelated incidents
Deepfake Obama Introduction of Deepfakes
· 29 reports
Defamation via AutoComplete
· 28 reports