P2ACH AI Vision AI products may analyze camera or CCTV video inputs from customer-operated stores, retail spaces, DOOH media, hotels, transport or mobility environments. Purposes include aggregated and statistical insights such as visitor counts, foot traffic, paths, congestion, ad attention, gaze dwell time, estimated age group and gender, Re-ID-based deduplication and heatmaps. Regardless of whether original footage is stored, video frames may be treated as personal information or video personal information during processing under applicable law, so the customer site legal basis, notices, rights process and contract/DPA apply together.
Fixed video-device operation and management matters- Legal basis and purpose: deployments are based on applicable law, customer contracts, PoC notices or other lawful authority, and are limited to statistics and operational insights.
- Number, location and filming range: where P2ACH AI directly installs or operates devices, the actual number, location and filming range are disclosed through the on-site notice, contract, DPA or separate notice. For customer-operated sites, the customer notice and policy control.
- Manager, team and access rights: the personal video information manager is Dongyul Lee, CEO; the responsible team is P2ACH AI Inc. Privacy Team; access is limited to personnel with a contract, operations or security need.
- Filming time, retention, storage and processing: filming time follows each site setting. As a product principle, we do not store original CCTV video or image frames. Frames are analyzed in memory on Edge AI devices; video information is pseudonymized and then deleted immediately; only aggregated, statistical or de-identified category data and event results are generated and transmitted.
- Access and rights requests: requests to view, confirm, delete or suspend processing are handled through the site operator, contract contact or Privacy Team, with identity verification and masking or restriction where third-party rights may be affected.
- Safeguards and other notices: safeguards include least-privilege access, access logs, TLS, Edge device and admin access controls, staff training, processor management and physical access control. On-site notices should state the purpose/place, filming range/time, manager name/title/contact and, if outsourced, processor name/contact. Audio recording, arbitrary camera manipulation and filming in privacy-sensitive places are not permitted.
We do not use Vision AI outputs for face-recognition identity verification, matching with names or contact details, biometric authentication or tracking intended to identify a specific person. Depending on the customer contract and actual role allocation, P2ACH AI may act as a processor/service provider or as a joint controller for customer-site Vision AI processing.
If video AI outputs are handled under pseudonymized-data rules, we separate or delete additional information where it is not needed, separate access rights for pseudonymized data and additional information, maintain processing records, prohibit re-identification attempts, apply least-privilege access and logging, impose contractual no-reidentification duties and perform periodic checks. For Korea, customer on-site signage, contracts or DPAs address installation purpose limits, filming range/time, person in charge, no audio recording, no arbitrary camera manipulation, no use beyond stated purposes and no filming in privacy-sensitive areas. For the EU/EEA, we consider EDPB video-device guidance on lawful basis, transparency notices, minimization, purpose limitation, retention limitation and data-subject rights. For Japan, we consider PPC guidance on camera images and facial-feature data, including purpose specification, retention, outsourcing/joint-use structure and absence of identification purpose. For Vietnam, we consider Decree 13 notice, consent and data-subject-right principles and public-place video/audio recording notice principles. For the United States/California, we consider CCPA/CPRA notice, deletion, correction, opt-out and non-discrimination rights for video, inferred characteristics and information that may be perceived as biometric or sensitive personal information.