Abstract

When a user searches “plumbers near me” on a map or marketplace and contacts a few businesses, those businesses gain a contact reference that keeps working after the inquiry ends: a phone number, masked number, chat thread or callback link. Masked and virtual numbers hide the real number but remain reachable paths. Simply possessing the reference is enough to call or message again. The results include:

  • follow-up spam and robocalls;
  • resale of lead data;
  • contact across unrelated inquiries;
  • scraper-driven abuse;
  • users who avoid contacting businesses at all.

AI agents now make this worse. They call and message businesses on users’ behalf, and business-side AI agents call users back, both potentially at machine scale. Operators fight spam by detection: analytics score and label calls, and caller-ID authentication such as STIR/SHAKEN confirms who is calling. Neither answers whether this caller is allowed to reach this person, for this purpose, now.

This disclosure describes a pre-effectuation communication-control architecture:

  • Non-bearer contact references. Every phone number, masked number, chat handle, callback link or AI-agent contact reference is non-bearer. Knowing, storing, reselling or replaying it does not allow contact.
  • Query-scoped handles. Each inquiry receives its own handle.
  • Verification at an enforcement point.A business or AI agent can contact the user only through an enforcement point that verifies a signed, context-bound authorization. Enforcement points include:
    • session border controllers, SIP proxies and IMS cores;
    • WebRTC and TURN servers;
    • SMS, RCS and email gateways;
    • push or device brokers;
    • CPaaS gateways and marketplace messaging servers.
  • What the authorization is bound to.It covers:
    • the handle, the inquiry and the business identity;
    • the requesting actor, human or AI agent;
    • the channel and the permitted effect;
    • validity, quota and revocation conditions.
  • No path until verified. Until verification succeeds, no call bridge, SIP dialog, media path, chat thread, message delivery or notification exists. The path is absent rather than blocked.
  • Preview, then separate future authorization. A short, bounded preview lets the user compare businesses. Further contact requires a separate future authorization, issued only after the user selects, approves or books. The business may pay a lead or callback-extension fee, but it never learns the user’s permanent number.
  • No reuse or transfer. Authority for one inquiry or business cannot be reused for another. It cannot be transferred to affiliates, resellers or call centres, or widened through AI sub-agents. API credentials, OAuth sessions, platform accounts or carrier session setup alone do not authorize contact.
  • Fast enough for the call path. A canonical scope descriptor is hashed in advance. The enforcement point compares an attempt hash and atomically updates nonce, quota and revocation state in sub-millisecond to low-millisecond time. It then issues a short-lived bounded release capability, which the media or messaging gateway consumes.
  • Other protections. Keys and state can sit in a TEE, HSM, secure element or modem security processor. User revocation ends contact instantly. Signed receipts provide evidence. Failures fail closed with no fallback path.

The claims run to claim 139 and include seven independent claims (1, 2, 79, 79A, 136, 138 and 139). They cover:

  • enforcement-point methods and systems;
  • privacy-preserved lead monetization;
  • descriptor-hash hot-path verification;
  • device-side release brokers;
  • bounded release capabilities;
  • AI-agent-initiated and business-side AI-agent contact;
  • recursive delegation tokens;
  • proactive revocation;
  • cross-service purpose control.

Industry relevance (AI-generated). This section was generated with AI assistance and describes technical relevance only. Readers must cross-check it against the original PCT filing on WIPO PATENTSCOPE and official company sources before drawing any conclusion. Naming a company or product is not an endorsement, a validation, a criticism or a statement of infringement, and no company named is affiliated with the inventor. These are potential integration scenarios, not confirmed company implementations or product commitments.

The architecture complements detection-based anti-spam with authorization-based admission in the call path:

  • Operator anti-spam: Verizon Call Filter, AT&T ActiveArmor and Airtel’s network-based AI spam detection.
  • Caller authentication: STIR/SHAKEN and India’s DLT framework.
  • Network APIs: Aduna, the joint venture of Ericsson with operators including AT&T, Verizon, T-Mobile, Vodafone, Orange, Deutsche Telekom, Telefónica, Jio and Airtel, which offers CAMARA-based APIs such as Number Verification and SIM Swap. A “verify contact authority” API fits naturally beside them.

For AI-native 5G-Advanced and 6G, it aligns with:

  • Verizon: the 6G Innovation Forum with Ericsson, Nokia, Samsung, Meta, Qualcomm and NVIDIA.
  • NVIDIA and Nokia: the AI-RAN partnership on the ARC-Pro platform.
  • Huawei: AgenticCore and the A2A-T agent protocol.
  • Qualcomm: the planned 2028 pre-commercial 6G devices.
  • O-RAN: RAN Intelligent Controller (RIC) applications.

Timing with telecom roadmaps (AI-generated). This paragraph was generated with AI assistance from public company sources. Readers must cross-check it against the original PCT filing on WIPO PATENTSCOPE and official company sources before drawing any conclusion. It is not an endorsement, a validation, a criticism or a statement of infringement, and it describes potential integration scenarios, not company plans or commitments.

The disclosure arrives as the industry moves from 5G-Advanced toward AI-native 6G. 3GPP is studying 6G in Release 20, with first 6G specifications targeted for Release 21 around the end of 2028. Operator 6G labs and trials are planned toward 2028, and commercial 6G is expected around 2030. Across this 2026–2030 window, operators and vendors are deciding:

  • how AI agents will place calls and send messages on networks;
  • which functions run at the AI-RAN edge;
  • which capabilities are exposed as network APIs.

Three current programmes map directly onto the claims:

  • Orange (Orange LiveNet and Orange OpenTech).Orange LiveNet is Orange’s network-API business unit, launched in February 2025. Its first commercial APIs are identity and anti-fraud APIs, offered through developer.orange.com and Aduna, and it is preparing APIs for agentic AI. Orange OpenTech, Orange’s annual innovation event, has featured network-API conferences and hackathons; its 2026 edition is scheduled for 17–19 November 2026. A “verify contact authority” API would sit naturally next to Orange LiveNet’s identity and anti-fraud APIs. The relevant claims are:
    • the enforcement point (claims 1 and 2);
    • privacy-preserved lead monetization offered to platforms (claims 79 and 79A);
    • descriptor-hash verification inside the API path (claim 136);
    • user-initiated revocation (claims 60 and 61).
  • Nokia (Network as Code, IMS and voice cores, AirScale baseband with anyRAN).The check can run where the network already sits. The relevant claims are:
    • verification sold as a developer API on Network as Code (claims 79 and 136);
    • enforcement in IMS cores and session border controllers before any SIP dialog exists (claims 1, 2 and 99–103);
    • a bounded release capability consumed at the gateway (claim 139);
    • deployment on anyRAN-based cloud or purpose-built RAN.
  • NVIDIA (AI Aerial and AI-RAN on the ARC-Pro platform).The hot path can run at the edge, next to voice-AI workloads. The relevant claims are:
    • hot-path verification at the edge (claims 13–15, 57 and 58);
    • AI-agent-initiated contact (claims 41, 42 and 104–106);
    • edge analytics that drive proactive revocation (claims 123 and 124).

Together, these place query-scoped, non-bearer communication authority in the layers now being designed for AI-native networks: the network API, the core and SBC, and the AI-RAN edge.

Further relevance:

  • CPaaS: Twilio Programmable Voice, masked sessions, Voice Integrity, Branded Calling and ConversationRelay voice AI.
  • Marketplace messaging: Booking.com.
  • Map and local platforms: Google Maps and Business Profile, Apple Maps and Business Connect.
  • AI calling: Google’s agentic calling in Search, ChatGPT agent, Claude and Copilot.

New revenue can come from:

  • verified-reach APIs priced per verification;
  • consumer “no post-inquiry contact” tiers;
  • callback-extension fees paid by businesses instead of buying phone numbers;
  • higher connect rates for authorized business calls;
  • lower fraud and complaint costs.

The mechanisms may support GDPR purpose limitation and consent withdrawal, ePrivacy rules on unsolicited communications, EU AI Act transparency for AI interactions, the U.S. TCPA, and India’s TRAI anti-spam regulations. Technical alignment does not itself establish regulatory compliance.

This disclosure is based on PCT/IB2026/056771 (WO 2026/172333 A2, filed 1 July 2026, published 20 August 2026), which claims priority to PCT/IB2025/058316. It belongs to the family built on the Mothership application PCT/IB2026/055615, filed 4 June 2026 and publicly disclosed 16 July 2026. The inventor is open to acquihire or strategic collaboration.

Keywords: AI safety; AI agent calling; AI voice agents; agentic communication; anti-spam; anti-robocall; anti-harassment; spam call prevention; robocall mitigation; caller authorization; STIR/SHAKEN complement; branded calling; number masking; proxy numbers; virtual numbers; masked calling; privacy-preserving contact; consent-based communication; purpose-bound communication; lead privacy; lead monetization; pay-per-lead; callback control; map business discovery; local search; marketplace messaging; host-guest messaging; click-to-call; non-bearer token; capability token; context-bound authorization; anti-replay; quota; revocation; SIP; IMS; session border controller; WebRTC; TURN; SMS; RCS; CPaaS; network APIs; CAMARA; GSMA Open Gateway; Aduna; Number Verification; Orange LiveNet; Orange OpenTech; developer.orange.com; Network as Code; anyRAN; AirScale; AI Aerial; ARC-Pro; 3GPP Release 20; Release 21; IMT-2030; 5G-Advanced; 6G; AI-native networks; AI-RAN; O-RAN; RIC; edge computing; TEE; HSM; fail-closed; GDPR; ePrivacy; TCPA; TRAI; Verizon; AT&T; Vodafone; Orange; Jio; Airtel; Nokia; Ericsson; Huawei; NVIDIA; Qualcomm; Twilio; Booking.com; Google Maps; Gemini; ChatGPT; Claude; Copilot

Creative Commons License

Creative Commons License
This work is licensed under a Creative Commons Attribution-Noncommercial 4.0 License

Download

Share

COinS