SafeBind.ai
FAQ
What publishers, buyers, and their counsel ask us most — about pricing, independent verification, and how SafeBind actually works.
SafeBind.ai reduces TCPA litigation risk for both lead publishers and lead buyers. We do it in five ways, all from the same capture and all in the same save.
They divide into two halves, deliberately. The Evidence is the record — the facts of the consent moment, signed and timestamped, and nothing else. Screening is the analysis — the judgments we make about the lead, returned by the buyer call on the record URL and never written into The Evidence. (A few aggregate analyses, like seller-side screening and cross-lead reuse, stay in your console rather than the API.) A record that grades its own consent hands the other side an exhibit that argues with itself; one that simply states what happened does not.
1 · Consent verification. A recording of the consumer filling out the form and signing the TCPA consent language as it was presented — the disclosure verbatim, its rendered font size and contrast ratio, whether the consent control was pre-checked when the page loaded, and the specific opt-in action taken. Signed into The Evidence.
2 · Named Buyer verification. Confirmation for the buyer that the consumer explicitly agreed to be contacted in a manner that buyer approved. Returned by the API; the parties named in the disclosure are signed in The Evidence.
3 · Consent channel verification. Analysis of the exact consent language to confirm it explicitly authorizes the buyer’s required communication methods — SMS, autodialer, automated voice, AI voice and calls from a person. Returned by the API; the language it analyzes is signed in The Evidence.
4 · Data integrity verification. Screening for markers of synthetic or falsified consumer information, such as gibberish names, disposable email addresses, or suspicious scripted form-fill behavior. Returned by the API.
5 · Network and fraud signal detection. Analysis of the lead’s technical origin for anomalies that suggest non-human or masked traffic — bot signatures, spoofed devices (a browser whose graphics chip, screen or features contradict the device it claims to be), a device that changes partway through one visit, VPN or datacenter usage (VPN through your own MaxMind key), timezone mismatches, unverified capture domains, and IP addresses geographically disconnected from the stated address. Returned by the API.
The Evidence is signed (ES256), independently timestamped (RFC-3161) and fingerprinted (SHA-256 + Merkle root), so anyone can verify it without trusting SafeBind. The competition is a hosted “trust us” legacy certificate with stacked per-certificate charges. SafeBind adds the checks they lack — one-to-one consent, the signed T&C / privacy documents, the consent control’s state on page load — with every risk data point in one pay-as-you-go save at a fraction of TrustedForm’s price. See the full comparison →
TrustedForm Verify confirms that consent language you pre-approved was present during the lead event: you manage a library of acceptable wording variations in their consent-language manager, and leads are automatically approved or rejected against those stored variations at acquisition (per their published documentation). Claiming a record separately offers a simpler required/forbidden-text check, included in the claim fee. What their published product doesn’t do is interpret wording it hasn’t been given — it can’t tell you whether an unfamiliar disclosure granted SMS or autodialer consent.
SafeBind runs both passes on every lead, in the same save. First the exact match: your fixed legal wording, clause-by-clause, with {phone} / {name} inserts verified against the consumer’s own details. Then the AI analysis: does the consent cover how you contact them, and is your company named? Every finding quotes the consent sentence it rests on. So “I need consent that permits SMS and autodialers” is a rule you set once, not a wording library you maintain. On their side, Verify is a separate 8–15 cents per-lead charge. See the line-by-line comparison →
No. Every record is signed with a key whose public half is published (JWKS). Anyone can check a record’s signature, The Evidence and its timestamp on their own machine, with no SafeBind server in the loop. One thing needs our published key list (JWKS): confirming the signing key is ours — fetch it once and keep it. More on security →
Every legal export carries a receipt SafeBind signs. It states when the record was captured and signed, who the copy was provided to, and, for a lead buyer, the day that buyer saved it. The receipt is signed separately from The Evidence, so the record verifies on its own whether or not a receipt is attached. Change the name on a copy and the receipt stops verifying; The Evidence still does. Every page of the PDF carries the same stamp, and it always leads with the capture date, so a later save date is never mistaken for the date of the evidence.
It verifies: the signature and the timestamp are the same on every copy. But it is the lead seller’s copy. Its receipt names the lead seller, and the partner list page shows only its fingerprint. Save the record yourself to get the full partner list and a receipt in your company’s name, with the record kept for five years. Under our Terms, an export is for the account it was provided to: a lead seller shares a record with a lead buyer through SafeBind, not by forwarding its own copy. Either side may still give its copy to its counsel, insurers, regulators and courts.
The same Evidence as every other copy, signed and timestamped when the consumer consented, and the stamp leads with that date. Your receipt adds the day you saved it, because that is when SafeBind began holding it for you. Nothing about when the evidence was captured changes.
The partner list names the companies a consumer agreed to hear from. On records captured from September 19, 2026, your console and your exports show its fingerprint rather than the page, so you can still show it is the same capture. A lead buyer who saves the record receives the full list. If you need to see it on one of your records, contact us.
A fraction of TrustedForm’s price at every volume tier. A publisher saves its own record for 3–6 cents and a buyer saves one it bought for 4–7.5 cents — graduated tiers that step down as your saves in a billing period grow, with the first 1,000 at the top rate and the next 19,000 at the next. That is half of TrustedForm’s published Retain rate at the same tier (September 2026) — and Retain is only their first charge: Verify, five years of storage and the Insights fees (5 cents per data point; bot detection an upcharge, price unknown) stack on top of it there, while every save here includes the consent analysis, the fraud check and every risk data point, with the evidence kept for five years and unlimited free views for that full retention period. Publishers create records free and can view them for up to 90 days. Pay-as-you-go, with no contracts and no minimums — every tier is on the comparison page.
Each save is priced at the tier its position in your billing period falls in, and earlier saves are never re-rated. A publisher’s first 1,000 saves in a period are 6 cents each, the next 19,000 are 5 cents, the next 30,000 are 4 cents, and everything past 50,000 is 3 cents. A buyer’s price starts at 7.5 cents and steps down the same way. Publisher saves and buyer saves count separately. Your billing period is a month, starting on the day your account was created, and the count resets at each anniversary.
Because they are different saves. A publisher keeps a record captured on its own verified domain: evidence of its own work. A buyer saves a record captured on someone else’s domain, for a lead it bought. Each has its own volume tiers, and each list is half of TrustedForm’s corresponding published price list (September 2026).
Every session is available to the buyer and publisher within minutes of capture, signed and independently timestamped (RFC-3161) the moment they ask for it, with the exact consent instant recorded inside The Evidence — including a session where our detector never saw a consent submission. Detection is a machine’s reading of a page, not proof either way, so we sign regardless rather than decide on your behalf which evidence will turn out to matter.
The Evidence is then retained for 12 months from capture where a consent submission was detected, and for 90 days from capture where no consent submission was detected, whether anyone saved the record or not. Save one and it’s kept for five years. And if the free self-serve save window has already closed, the record still exists — it can be retrieved as a paid service for the rest of that retention period. You’re not punished for finding out late: where a consent submission was detected, a TCPA claim that lands in month four still has evidence behind it.
Add one JavaScript tag across your site, just like the tag you already run. Records are created immediately, and you can see a real lead scored against your own rules within minutes.
No — and we measure this rather than promise it. The tag is one small file (about 40 kB compressed — smaller than most images on your page) loaded from one server, in the background, so your page never waits on it. Visitors see your page just as fast, nothing shifts or jumps while it loads, and it doesn’t hurt the page-speed scores Google uses in ranking.
On our own production quote form, SafeBind’s total working time on the page measures around 20 milliseconds — about a fiftieth of a second, and a small fraction of what a typical analytics tool spends. While the consumer is filling out the form it stays out of the way: it records in tiny background batches, a few milliseconds at a time, and pauses itself when the visitor goes idle.
Paste your language with variable inserts where the per-lead details go — {phone}, {first_name}, {last_name}, {email}. SafeBind matches the fixed legal wording verbatim, then verifies each insert separately against the consumer’s own submitted contact. If the disclosure names a phone or email that conflicts with the consumer’s, the check fails and the reason comes back with the verification. Company opt-out or toll-free numbers in the disclosure are ignored, so they never cause a false failure.
No. SafeBind gives you strong evidence and automated consent checks, but whether your practices comply with the TCPA and privacy laws is your responsibility (and your counsel’s). See our Terms.