Your AI voice agent can hold a natural conversation, handle objections and book a meeting — but none of that matters if the call connects and the prospect’s screen says “Unknown” or, worse, “Spam Likely”. The caller-ID screen is the first impression of every outbound call, and for most legitimate businesses it is an impression they do not control. Branded calling is the set of technologies and commercial programmes that change that: your verified business name, logo and even a call reason on the recipient’s screen before they decide whether to answer.
This post is the practical deep-dive we promised in our STIR/SHAKEN explainer: what Rich Call Data actually is at the standards level, how branded calling is really delivered in the United States today, what exists (and does not exist) in Australia, and the honest limits of the whole approach.
Quick answer: Branded calling puts a verified business name, logo and call reason on the recipient’s caller-ID screen. In the US it is delivered two ways: vendor programmes (such as First Orion and Hiya) with carrier and device distribution deals, and the standards-based route — Rich Call Data (RCD), defined in RFC 9795 and RFC 9796 (July 2025), which cryptographically ties display name, logo and call reason to a STIR/SHAKEN-signed call. CTIA’s Branded Calling ID is the industry ecosystem built on those standards. Australia has neither STIR/SHAKEN nor an RCD ecosystem, so AU teams rely on caller-ID hygiene instead: consistent CLI, lawful overstamping and number reputation.
Zian’s outbound agents live or die on answer rates like every other calling operation, which is why we treat caller-ID presentation as infrastructure, not an afterthought. If you want AI agents that take the phone channel seriously end to end, Apply For Partnership.
The problem: legitimate calls that look like spam
When your outbound number has no branding, the recipient’s carrier and handset decide what to display. That usually means a bare number, a generic location string, or an analytics-driven warning label such as “Spam Likely” or “Scam Likely” if the number’s calling pattern has tripped a reputation model. High-volume outbound — exactly what an AI cold-calling operation produces — is precisely the traffic shape those models watch.
We will keep the impact claims qualitative, because most published “branded calls lift answer rates by X%” figures come from vendors selling branded calling and are not independently verifiable. What is uncontroversial: people are more likely to answer a call when the screen shows a name they recognise and a reason for the call, and unlabelled unknown numbers are widely ignored. If your AI agent’s first-second experience matters, the pre-answer screen matters even more — it decides whether there is a first second at all.
Rich Call Data: the standards layer
Branded calling in the US is converging on a standards stack called Rich Call Data (RCD), which extends the STIR/SHAKEN caller-ID authentication framework. The base layer is the SIP identity architecture of RFC 8224 and the PASSporT signed token of RFC 8225 — the machinery that lets a carrier cryptographically attest that a call is authorised to use its calling number. (Our STIR/SHAKEN post covers attestation levels and what they mean for AI voice deployments.)
In July 2025 the IETF published two Proposed Standards that define how branding rides on top of that machinery — and it is worth being precise about which document does what:
- RFC 9795 — “Personal Assertion Token (PASSporT) Extension for Rich Call Data” extends the signed PASSporT token itself. It defines the claims that carry the branding payload:
rcd(the rich call data — display name, icons/logos, jCard contact data),rcdi(integrity protection for that data) andcrn(the call reason). In the RFC’s own words, it extends PASSporT “to include rich metadata about a call and caller that can be signed and integrity protected, transmitted, and subsequently rendered to the called party”. - RFC 9796 — “SIP Call-Info Parameters for Rich Call Data” defines how that information travels in SIP signalling: new Call-Info header parameters (
call-reason,verified,integrity) and ajcardpurpose token, aligned with the RCD PASSporT.
The design goal is the important part for buyers: the name, logo and call reason are cryptographically tied to the same signed identity assertion that STIR/SHAKEN uses. A terminating carrier or device that supports RCD can verify that the branding was asserted by an authorised signer and has not been tampered with in transit — which is what separates standards-based branded calling from simply asking a caller-ID app to show a nicer name.
How branded calling is actually delivered in the US today
Standards describe the destination; delivery today is a mix of commercial programmes. Two things are true at once: vendor programmes with carrier distribution deals are what most enterprises actually buy right now, and the standards-based ecosystem is being stood up around them.
Vendor programmes
First Orion sells branded calling under its INFORM product family. Per First Orion’s own product pages, INFORM displays a business name of up to 32 characters on outbound calls, INFORM with Logo adds a logo and call reason (delivered via API or SIP integration), and the company says it delivers branded calls across major US carriers including T-Mobile, AT&T, Verizon and Boost Mobile, reaching nearly 400 million phones in the US. Those reach figures are First Orion’s claims, verified on First Orion’s site — treat them as vendor-stated, not independently audited.
Hiya offers branded call capability that, per hiya.com, lets a business show “your name, logo, and reason for calling” on outbound calls. Hiya’s distribution rides on its caller-ID app and device/carrier partnerships.
The industry ecosystem: CTIA’s Branded Calling ID
Branded Calling ID (BCID) is the US wireless industry’s standards-based programme, operated by CTIA. Per brandedcallingid.com, BCID describes itself as “the only industry-led, standards-based Rich Call Data (RCD) ecosystem engineered to be secure-by-design”, conveying name, logo and call reason over “in-band, end-to-end cryptographically signed calls” leveraging STIR/SHAKEN — i.e., the RFC 9795/9796 approach operationalised, with vetting of enterprises before their brand can be displayed and delivery confirmation back to the enterprise. In a 15 September 2025 press release, CTIA announced a signed contract to integrate BCID on Verizon’s wireless network — a genuine milestone, though the release did not commit to a consumer availability date.
A verification note, because honesty is the point of this blog: everything above was checked at the programme owners’ or vendors’ own sites. We deliberately have not named other branded-calling resellers whose claims we could only find repeated on marketing blogs. If you are evaluating a provider, ask which delivery mechanism sits underneath — an app install base, a direct carrier deal, or signed RCD — because that determines where your logo can actually appear.
Zian deploys voice agents on properly provisioned numbers with the carrier plumbing sorted per market — that is table stakes for the answer rates our platform is optimised around. Apply For Partnership to see how we run the phone channel.
What the recipient sees: four approaches compared
| Approach | What the recipient sees | Who controls it | Availability |
|---|---|---|---|
| Raw CLI (no branding) | Bare number, carrier-guessed name, or an analytics label like “Spam Likely” | Terminating carrier and analytics apps — not you | Default everywhere |
| STIR/SHAKEN attestation only | Often a “verified” indicator or checkmark at best; no name or logo | Originating carrier signs; terminating carrier decides display | US and some other markets; mandated for US IP networks |
| Branded calling / RCD | Business name, logo and call reason, cryptographically tied to the signed call (standards route) or delivered via vendor/app programmes | You (via a vetted programme), within carrier and handset support | US: live via vendors; BCID ecosystem rolling out |
| Australian status quo | Number and, where supported, calling-number display; no verified logo/name ecosystem | Your CLI choices plus telco scam-filtering under C661:2022 | AU today — no STIR/SHAKEN, no RCD |
Australia: no RCD ecosystem — what AU teams can do instead
Australia has not adopted STIR/SHAKEN, and there is no Australian Rich Call Data or branded-calling ecosystem: no framework exists for a verified logo or brand name to be cryptographically delivered to Australian handsets. The Australian framework is oriented at blocking bad calls rather than authenticating good ones. The ACMA states it has registered Industry Code C661:2022 Reducing Scam Calls and Scam SMs, which “requires telcos to identify, trace and block SMS scams and scam calls” — and telcos reported blocking more than 2.2 billion scam calls since December 2020 (ACMA, data current as at November 2024).
That framework still shapes what a legitimate AI-calling operation should do in Australia:
- Use consistent, real CLI. Present a stable Australian number that connects back to you. Per the same ACMA page, CLI overstamping — displaying a different number than the one you are calling from — is legal in Australia for legitimate purposes, but freephone and local-rate numbers (13, 1300, 1800) and invalid or unallocated Australian numbers cannot be used for it. Overstamp only to numbers you hold, and discuss CLI management with your carrier.
- Provision properly. Number type, carrier relationships and CLI rules for AU voice agents are covered in our +61 provisioning guide.
- Protect number reputation. Under C661-era scam filtering, patterns matter: sudden volume spikes, high short-call ratios and unanswered-call churn from a single CLI look like the traffic telcos are required to disrupt. Warm numbers up, spread volume sensibly, and keep answer-driven pacing.
- Brand the channels that do support it. Australia’s verified-sender machinery currently lives in SMS, not voice — see our guide to the SMS Sender ID Register, where registered alphanumeric sender IDs are real and enforceable today.
Honest limits: what branded calling does not do
Branded calling is presentation-layer trust. It does not make an unlawful call lawful: consent, calling-hour rules, the Do Not Call Register in Australia, TCPA obligations in the US, and AI disclosure on the call itself are entirely separate obligations that a logo on the screen does nothing to satisfy. A branded spam call is still a spam call — and a fast way to get a vetted brand identity revoked.
Display behaviour also varies. Whether your name and logo actually render depends on the terminating carrier’s support, the recipient’s handset and OS, and (for app-based programmes) whether the relevant app is installed. STIR/SHAKEN attestation itself does not carry your logo anywhere — the SHAKEN attest claim (RFC 8588) has values A, B or C for full, partial and gateway attestation of the calling number, and even A-level attestation only vouches for number authority, not branding. And no programme guarantees immunity from analytics labels: reputation models are run by carriers and their analytics partners on many signals, of which branding participation is one.
Where this fits for AI voice agents
For AI-driven outbound, branded calling closes a loop the conversation layer cannot. An agent with sub-second latency and a good script never gets to use either if the call is screened out at the glass. In the US, that argues for putting your programme choice (vendor-delivered vs standards-based RCD) on the same evaluation sheet as your dialling infrastructure. In Australia, it argues for disciplined CLI and reputation hygiene while the ecosystem catches up — and for watching whether the RCD standards eventually travel here, as we flagged in the STIR/SHAKEN comparison.
FAQs
What is Rich Call Data?
Rich Call Data (RCD) is caller information beyond the phone number — display name, logo or icon, contact card and a call reason — carried with a call in a cryptographically signed form. It is defined in RFC 9795 (July 2025), which extends the PASSporT token used by STIR/SHAKEN so that rich metadata about the caller “can be signed and integrity protected, transmitted, and subsequently rendered to the called party”, with companion SIP signalling defined in RFC 9796.
Does branded calling stop “Spam Likely” labels?
Not by itself. Spam and scam labels are applied by terminating carriers and their analytics partners based on many signals — calling patterns, complaint data, number history — and branded-calling participation is one signal among several. A vetted brand identity helps, but poor calling behaviour on a branded number can still earn labels, and labelling policies differ by carrier and app.
Is branded calling available in Australia?
No. Australia has no STIR/SHAKEN deployment and no Rich Call Data ecosystem, so there is no way to deliver a verified name and logo to Australian handsets today. Australia’s regime targets scam disruption instead: the ACMA has registered Industry Code C661:2022 Reducing Scam Calls and Scam SMs, requiring telcos to identify, trace and block scam calls and SMS. Australian teams should focus on consistent lawful CLI, number reputation and the SMS Sender ID Register for branded SMS.
Does STIR/SHAKEN attestation guarantee my logo shows?
No. Attestation (the attest claim of RFC 8588, values A, B or C) only asserts the carrier’s confidence that you are authorised to use the calling number. Displaying a name or logo additionally requires an RCD or branded-calling programme, plus support at the terminating carrier and handset — and rendering still varies by device even inside supported programmes.
Who runs Branded Calling ID (BCID)?
BCID is operated by CTIA, the US wireless industry association, as a standards-based Rich Call Data ecosystem built on STIR/SHAKEN-signed calls, with enterprise vetting and delivery confirmation. Per CTIA’s September 2025 announcement, Verizon has contracted to integrate BCID on its wireless network. Details are at brandedcallingid.com.
Answer rates are won before the first word. If you want outbound AI agents built by people who sweat the caller-ID layer as hard as the conversation layer, Apply For Partnership.