Resources · Getting Recommended by AI

Which Credentials and Compliance Details Should SaaS Companies Publish for AI Trust?

SaaS companies should publish the credentials and compliance details a security reviewer would check, in text, on crawlable pages. List the certifications and attestations you hold, with their scope, dates and issuers. Say where data is stored and processed, list your subprocessors, and explain how buyers can get the full reports. Link to public registries where your status can be verified. A SOC 2 report is a restricted-use report, so publish a clear summary, or a SOC 3 report, which is meant for general use. Assistants asked "is [vendor] SOC 2 compliant?" or "which [category] vendors keep data in the EU?" can only answer from what is public. Vague claims such as "enterprise-grade security" give them nothing to repeat.

What should a trust centre contain?

Section What to publish
Summary One paragraph: the certifications and attestations you hold, hosting regions, and how to request reports
Certifications and attestations Each one with its name, scope, issuing body or auditor, date issued or period covered, and renewal cycle
Verification links Public registry entries where they exist
Data location Hosting regions, where backups are kept, where support staff can access data from
Subprocessors Name, purpose and location of each, with a change log
Data protection Encryption at rest and in transit, key management, retention and deletion
Access and identity Single sign-on, multi-factor authentication, role-based access, audit logs
Incident and vulnerability handling How to report a vulnerability, response commitments, status page
Legal documents Data processing agreement, transfer mechanisms, privacy policy
Reports on request How to get the SOC 2 report, penetration test summary, and questionnaire answers

Keep it in plain HTML, not behind a login or loaded by script. Buyers can still request restricted documents through a form. The summary facts must be public for assistants to use them.

Which pages should a SaaS trust center publish to pass AI-assisted security reviews?

Security reviewers increasingly use AI tools to pre-fill questionnaires and summarise vendors. Those tools, and the assistants buyers use, look for:

  • A certifications page with scope and dates, not just badges.
  • A subprocessor list with locations.
  • A data residency page, covered below.
  • A security overview answering the common questionnaire topics in prose: encryption, access control, logging, backups, business continuity, secure development.
  • A vulnerability disclosure page.
  • A status and incident history page.
  • Answers to standard questionnaires, summarised publicly, with full versions on request.

Each page should state facts that can be checked, with dates. "Encrypted at rest with AES-256; keys managed in a cloud key management service; rotated annually" is useful. "Bank-level security" is not.

How should a security SaaS explain SOC 2 Type II evidence for AI retrieval?

The American Institute of CPAs describes SOC 2 as an examination of controls relevant to security, availability, processing integrity, confidentiality or privacy. A Type II report covers how controls operated over a period, commonly three to twelve months, rather than at a single point in time. The full report is restricted-use. The AICPA describes SOC 3 reports as general-use reports that can be freely distributed. So publish:

  • The report type and period: "SOC 2 Type II, covering 1 January to 31 December 2025."
  • The criteria in scope: for example, security and availability.
  • The systems in scope: which products and environments.
  • The auditor: the CPA firm that performed the examination.
  • The opinion: whether it was unqualified, stated accurately.
  • A SOC 3 report or a summary for general reading, and how to request the full SOC 2 under a confidentiality agreement.
  • The next report date.

Avoid "SOC 2 compliant" as a bare claim. SOC 2 is an attestation for a period and scope, not a certificate. Assistants repeating a precise statement are more useful to buyers than ones repeating a slogan.

What data residency evidence do AI assistants need before recommending a B2B SaaS vendor?

Questions such as "project management tools that store data in the EU" filter vendors, so the evidence must be explicit:

  • Regions offered, and whether customers can choose, by plan.
  • What stays in region: primary data, backups, logs, search indexes, AI features.
  • What may leave: support access, subprocessors outside the region, and the safeguards used.
  • Transfer mechanisms: standard contractual clauses, and certification under the EU-US Data Privacy Framework where held. The US Department of Commerce publishes a searchable list of participants, so link to your entry.
  • Subprocessors by location.

Put it on a dedicated data residency page and repeat the headline fact on pricing and product pages, such as "EU data residency available on Business and Enterprise plans".

What evidence should an identity verification vendor publish to earn recommendations from AI assistants?

Identity verification buyers ask about accuracy, fraud resistance and regulatory fit. Publish:

  • Independent testing: for example, presentation attack detection testing against ISO/IEC 30107-3 by an accredited laboratory, with the confirmation letter and test level.
  • Certifications and registrations, such as inclusion on the UK government's register of digital verification services certified against its trust framework, or equivalent schemes in your markets.
  • Supported documents and countries, as a searchable list.
  • Performance measures you can support, with how they were measured.
  • Data handling for biometrics: retention, deletion and consent.
  • Regulatory use cases: which know-your-customer, anti-money-laundering or age verification requirements customers use you for.

Link each claim to its source. Identity is a field where buyers, and assistants, distrust unverifiable claims, and where an overstated accuracy figure can become a compliance problem for customers who relied on it.

How do you check what assistants say?

Ask ChatGPT, Gemini, Perplexity and Claude: "Is [vendor] SOC 2 Type II?", "Does [vendor] offer EU data residency?", "Is [vendor] ISO 27001 certified?", "[category] vendors with [certification]". Record whether answers are correct and current, and which sources they cite. Fix wrong answers at the source: your trust centre, review profiles, or outdated articles.

What might an illustrative case look like?

Consider an illustrative HR software vendor whose trust centre was a page of badges and a request form. Assistants said it was "not known to offer EU data residency" and gave its ISO 27001 status as "unclear". It rebuilt the trust centre in text, adding a data residency page listing regions by plan and a subprocessor list with locations. It published its SOC 2 Type II period, scope and auditor, and a SOC 3 report. It linked its ISO 27001 certificate to the certification body's public record. Two months later, most answers confirmed its residency options and certifications, citing the trust centre.

What do people ask about compliance pages and AI?

Should we publish our SOC 2 report?

No. It is restricted-use. Publish a SOC 3 report or an accurate summary, and share the full report under a confidentiality agreement.

Do badges help?

Only with text beside them. Assistants read the words, and buyers need the scope, dates and issuer.

Should we list certifications we are working towards?

Yes, if clearly labelled as in progress with an expected period, never presented as held. Assistants and buyers both need to tell the two apart.

How often should the trust centre be updated?

Whenever a certification renews, a subprocessor changes, or a region is added, with the date of each change recorded on the page.

Published by The Kovmere Team · updated Sep 29, 2026 · More resources