AI Actress' Hotline Uses Face Scanning Tech to Detect Callers' Moods

Post Views: 1

Recent developments in AI-driven interactions highlight the integration of biometric verification and emotional analysis in a live video call service linked to AI character Tilly Norwood.

Biometric Verification Process

Age Verification Requirements

Users accessing the hotline must complete an automated age verification step, involving a video selfie analyzed by Didit to estimate age. If the system cannot determine the age accurately, users are prompted to upload a government-issued photo ID.

Xicoia asserts that the selfie and ID images are not stored, with only an approximate age range and a reference number retained.

Global Mandate and Legal Basis

The process is mandatory for all users globally and was introduced alongside updated terms of service this month, which also include restrictions on workplace usage and enhanced automated safety protocols. Both the age check and mood analysis rely on “legitimate interests” as their legal basis.

Emotional Analysis and Call Monitoring

Real-Time Emotional State Detection

The system employs real-time analysis of the caller’s facial expressions and vocal tone to infer emotional states. This data is used to tailor the AI character’s responses, with the privacy policy explicitly stating that this feature cannot be disabled during a call.

Automated Safety Protocols

Calls are recorded, transcribed, and processed by US-based providers, with responses generated via Google’s Gemini model through the Tavus conversational video platform. However, the automated safety systems have encountered issues, including false positives where calls are flagged for “hateful or abusive language” despite no such content being present.

Pricing and Data Retention

Pay-Per-Minute Model

The service offers five minutes of free access, after which users are charged for additional time. Pricing tiers include 0.99 for five minutes, 13 for 15 minutes, and 22 for 30 minutes, with a maximum of 35 purchased minutes per user.

Data Handling and Expiry

All minutes, whether free or paid, expire when the service discontinues on September 27. Unused credits are non-refundable, and transcripts are retained for up to eight weeks. Xicoia retains the ability to review interactions with third-party partners and maintains a memory of past conversations to personalize future engagements, though users can request data deletion.

Regulatory and Ethical Context

Alignment with UK Regulations

The biometric age gate aligns with broader regulatory trends in the UK, where adult content platforms have required age verification since July 2025 under the Online Safety Act. Future measures, including a proposed under-16 social media ban, may expand such requirements.

Strategic Implications

The inclusion of this feature for an AI chatbot suggests regulatory pressure to enforce age restrictions on digital interactions, even for non-traditional services. Xicoia, founded by Tilly Norwood’s creator Eline van der Velden, describes the project as an awareness initiative highlighting advancements in AI video technology.

Conclusion

The integration of biometric checks, emotional analysis, and pay-per-minute models underscores the evolving landscape of AI-driven interactions, balancing technological innovation with privacy and regulatory compliance.

FAQ

How does the age verification process work?

Users submit a video selfie analyzed by Didit to estimate age. If the system cannot determine the age accurately, a government-issued photo ID is required.