Skip to content
AI Voice Agents
AI Voice Agents11 min read0 views

Carrier Filtering and Spam Scoring in 2026: How AI Voice Avoids the 'Spam Likely' Trap

Hiya, First Orion, TNS - three independent reputation databases, three different scores, one ugly outcome if you trip them. Here is how AI voice deployments stay green-lit in 2026 across AT&T, T-Mobile, and Verizon.

"Spam Likely" is the most expensive label in outbound voice in 2026. AT&T runs Hiya, T-Mobile runs First Orion, Verizon runs TNS Call Guardian - three separate models, three reputation databases, one outcome if you flunk any of them. The math: a single label drops answer rates by 60 to 80 percent overnight, and the only fixes are reputation rehab (slow) or burning the DID (expensive at scale).

Background

Carrier-level call labeling started rolling out in 2018-2019 in response to the FCC's TRACED Act and consumer complaints. By 2026, three vendors dominate the US labeling market: Hiya partners with AT&T, First Orion powers T-Mobile's Scam Shield, and TNS Call Guardian sits inside Verizon's Call Filter. Each runs its own behavioral model on call patterns and consumer feedback. They do not share scores.

Common triggers across all three in 2026: average call duration under 30 seconds (signals abandons or robocalls), volume spikes (>200 calls per day per DID for new numbers), neighbor-spoofing patterns (calls to area codes that match the calling DID at suspicious rates), high call abandonment rates, consumer complaints in the carrier's own app.

STIR/SHAKEN attestation is the table-stakes mitigation: A-attestation says the originating carrier verified the caller's right to use the number; B and C attestation are weaker. In 2025, 85 percent of Tier-1 inter-carrier traffic was signed; smaller providers signed only 17.5 percent. A non-A-attested call from an unknown number is a high spam score by default.

Hear it before you finish reading

Talk to a live CallSphere AI voice agent in your browser — 60 seconds, no signup.

Try Live Demo →

Steps and config

flowchart TD
    A[Outbound call placed] --> B[STIR/SHAKEN sign]
    B -->|A-attestation| C[Originating carrier delivers]
    C --> D{Terminating carrier?}
    D -->|AT&T| E[Hiya score lookup]
    D -->|T-Mobile| F[First Orion score lookup]
    D -->|Verizon| G[TNS score lookup]
    E --> H{Score above threshold?}
    F --> H
    G --> H
    H -->|Yes| I[Call delivered cleanly]
    H -->|No| J[Spam Likely label or block]
    J --> K[Reputation rehab needed]

Reputation rehab is slow. Hiya's free Voice Performance dashboard lets businesses see their score and dispute mislabels; First Orion offers similar through ENGAGE. Resolution timelines run 5 to 30 days.

CallSphere implementation

CallSphere monitors carrier reputation across our entire Twilio number pool nightly. Our number_reputation table (one of 115+ DB tables) tracks Hiya, First Orion, and TNS scores per DID, with quarantine triggers when any score crosses a configurable threshold. Sales Calling AI (1 of our 37 agents) checks the reputation cache before placing an outbound call and rotates DIDs if the matched local-presence number is degraded. STIR/SHAKEN A-attestation is automatic on Twilio-provisioned numbers. We submit reputation disputes through Hiya Connect's enterprise API on behalf of Scale ($1499/mo) tenants. Call duration analytics feed our Voice Insights dashboard so tenants can see where short-duration calls are eroding reputation. The 22% affiliate program credits Scale upgrades driven by reputation issues.

Build steps

  1. Provision DIDs from a single carrier with verified RespOrg status (Twilio is one); inherit A-attestation by default.
  2. Cap daily call volume per DID at 75 to start; raise gradually based on reputation telemetry.
  3. Track average call duration; calls under 15 seconds account for less than 30 percent of volume per DID.
  4. Pull reputation scores from Hiya, First Orion, and TNS nightly via their enterprise APIs or the Twilio Trust Hub.
  5. Quarantine DIDs that drop below threshold for 14 days; release after observed score recovery.
  6. Submit reputation disputes when a DID is mislabeled with documented call patterns.
  7. Enroll in branded calling (Hiya Connect, First Orion ENGAGE) to layer trust signals on top of clean reputation.
  8. Audit weekly: which DIDs answered, which got labeled, which had short-duration calls dragging the score.

FAQ

Why does my brand-new DID show as Spam Likely? Number reputation can carry over from previous owners after the FCC's 45-day aging period. Some carriers also apply a "new number" provisional score that takes weeks of clean traffic to clear.

How fast can I rehab a flagged DID? Best case: 5 to 10 days of clean traffic plus a successful dispute. Worst case: 30+ days, or quarantine and burn.

Still reading? Stop comparing — try CallSphere live.

CallSphere ships complete AI voice agents per industry — 14 tools for healthcare, 10 agents for real estate, 4 specialists for salons. See how it actually handles a call before you book a demo.

Does STIR/SHAKEN A-attestation prevent labeling? It helps but does not guarantee a clean delivery. A-attestation says the originating carrier verified you have the right to use the number; the terminating carrier's behavioral model still runs.

Can I see my Hiya score without paying? Hiya offers a free Voice Performance portal in 2026 that shows your score and basic call analytics. First Orion ENGAGE has a similar tier.

Is CallSphere's number pool safe by default? Yes for new tenants. Our quarantine and rotation logic prevents reputation cliff-drops. Tenants who upload their own DIDs without history may need a warm-up period.

Sources

Start a 14-day trial with managed reputation, browse pricing, or book a demo. Partners earn 22% via the affiliate program; enterprise reputation programs go to contact.

Share

Try CallSphere AI Voice Agents

See how AI voice agents work for your industry. Live demo available -- no signup required.

Related Articles You May Like

AI Voice Agents

WebRTC Mobile Testing with BrowserStack + Sauce Labs (2026)

BrowserStack offers 30,000+ real devices; Sauce Labs ships deep Appium automation. Here is how AI voice agent teams use both for WebRTC mobile QA in 2026.

AI Voice Agents

MOS Call Quality Scoring for AI Voice Operations in 2026: Beyond 4.2

MOS 4.3+ is the band where AI voice feels human. Drop below 3.6 and conversations break. Here is how to measure, improve, and alert on MOS in production AI voice using G.711, Opus, and the underlying packet loss / jitter / latency math.

AI Voice Agents

Claude-Powered Voice Agents for Salon and Spa Bookings

Why Claude salon AI is reshaping voice and chat automation, with concrete patterns for appointment AI in production deployments. A field-tested view from production teams shippi...

AI Voice Agents

Logistics Dispatch Voice Agent 2026: Driver Hotline + Load Assignment Hands-Free

Trucking dispatchers spend half their day on check-calls. Here is how a 2026 AI voice agent runs the driver hotline, assigns loads, and updates the TMS in real time.

AI Voice Agents

Build a CallSphere-Style Outbound Voice Campaign Tool

Replace expensive outbound SDR tooling with a self-hosted dialer that runs OpenAI Realtime agents at 100 concurrent calls. Full architecture and code.

AI Voice Agents

Claude for Real Estate Lead Routing and Follow-Up

Why Claude real estate is reshaping voice and chat automation, with concrete patterns for real estate AI in production deployments. A field-tested view from production teams shi...