Deep Labs Age Assurance

Case Study: Age Assurance

Massive user base
- • 2.5B monthly logged-in users
- • Top 3 global website & global app for traffic/use
- • 2x global user growth in last 10 years
- • 87% of US population uses site monthly
- • $8B in quarterly advertisement revenue
- • Youth content biggest driver in view counts
Increasing EU + US regulatory pressure & fines
- • Pressure to identify & screen underage users
- • EU + DSA1 threatening 6% of revenue as fine ($1B+)
- • 2x global user growth in last 10 years
- • Current FTC probe portends upwards of $2-3B+ in fines ($170M fine in 2019)
Brand and reputational damage necessitated immediate solution, without cannibalizing core business & disrupting existing customers
No organic automated ability to identify age
- • Limited capability via manually intensive processes
- • Limited ability to perform feature engineering for age assurance, fraud, and identity
- • Digital signatures for age assurance typically contaminated

• Large Signal Model (LSM) technology is regulatory compliant, with statistically significant outputs
• Industry leading platform at identifying actors/users pretending to be others
• AI infrastructure tuned to analogous digital signature challenges
• Deep experience aggregating holistic digital footprint signals & identifying anomalous behavior clusters

Improvement in identification of underage users
Number of human analysts, armed with DeepDecision庐 , required to produce output
Hours required to enable insights & analysis, identify anomalies, & generate report
Impact to existing customers, cannibalization of user base
Reduction in false positives from customer analysis (e.g. real users that aren鈥檛 underage)
Regulatory compliance, given statistically significant results from Large Signal Models (LSM)