STIR/SHAKEN and spam labeling: why your AI receptionist’s calls get flagged

You set up your AI receptionist to call customers back, confirm appointments, and follow up on quotes. Then customers start telling you your number shows up as “Spam Risk” on their phones. Your legitimate business calls are being treated like robocalls. This is the STIR/SHAKEN problem, and if you are buying an AI receptionist that makes outbound calls, you need to understand it before it costs you customers.

Why your calls get flagged

US carriers are required to run a caller ID authentication framework called STIR/SHAKEN on IP-based calls. In plain terms, the carrier that originates a call attaches a digital signature saying “we verified this caller.” The receiving carrier checks the signature and grades the call: full attestation means the carrier knows the caller and the number is legitimate, partial means the caller is known but the number use is unverified, and gateway means the call came from somewhere the carrier cannot verify at all.

When the signature is missing or weak, the receiving carrier may label the call “Spam Risk,” “Scam Likely,” or something similar. The scale of the underlying problem explains why carriers are aggressive: Americans received just under 3.9 billion robocalls in August 2026 alone, according to YouMail’s Robocall Index, and roughly half of those were telemarketing or scams. Your AI receptionist’s legitimate calls swim in the same water, so the carriers err on the side of flagging.

What full attestation requires

Full attestation is not automatic. The carrier originating the call has to verify two things: that you are who you say you are, and that you are authorized to use the number you are calling from. For a business, that usually means registering your business identity and your numbers with the carrier or through your AI receptionist vendor. If your vendor cannot tell you what attestation level your outbound calls get, that is a gap in their setup, not a mystery of the technology.

Questions to ask your AI receptionist vendor

First: what attestation level do my outbound calls receive, A, B, or C? You want A, full attestation, and you want it in writing. Second: are my business numbers registered, and with whom? Number registration through the proper channels is what separates your calls from spoofed ones in the carrier’s eyes. Third: what happens when a call gets flagged anyway? Some vendors monitor spam labeling and will tell you when your numbers pick up a bad reputation, and help you remediate. Others will not know until you complain.

Fourth, and most important: does the vendor handle this for you, or is it your job? The honest answer varies. Some AI receptionist platforms register your numbers and manage attestation as part of the service. Some hand you a carrier portal and wish you luck. Neither is wrong, but you need to know which one you are buying, because number registration done wrong is how legitimate businesses end up permanently labeled.

What you can do on your end

Keep your business listings consistent. Carriers and analytics engines cross-reference the number you call from against business directories, so a number that matches your Google Business Profile and your website is less likely to be flagged than one that appears nowhere. Do not rotate through new numbers frequently; number reputation builds over time, and a fresh number has no history to vouch for it. And watch your call patterns. Short calls to hundreds of numbers in an hour looks like a robocall campaign even when it is just aggressive follow-up. Pace outbound calling like a human would, because the filters are tuned for humans.

The inbound side matters too

Spam labeling cuts both ways. Your AI receptionist also needs to handle the flood of spam calls coming in, or your team will waste time on robocalls the AI should have filtered. Our spam-call filtering setup guide covers how to configure screening and block lists so the junk never reaches a human.