Security
Your customers' data. Handled carefully.
Two-factor sign-in, team roles, an activity log, PII scrubbing, retention you control, and approval rules on every AI action. Hosted on AWS, encrypted in transit.
Last reviewed Oct 24, 2026
Live
Live
Live
On our roadmap
How we protect your data
Four layers. Always on.
Sensitive details scrubbed from what we store. Consent and calling rules enforced on every call. Tight control over what the AI may do, and over what each person on your team may do.
Keep less. Expose nothing.
Aiva Chat scrubs personal details out of transcripts and summaries — card numbers, SSNs and dates of birth, plus any terms you add yourself. What you read in the dashboard is what you'd be comfortable keeping.
You decide how long conversation data is kept — retention is a setting on your account, not a fixed policy. Call recordings play back through short-lived links, so a copied URL stops working.
Callers always know. Opt-outs always stick.
Before any call is recorded, Aiva Chat plays a spoken recording notice. Texts honour STOP and START automatically, and anyone on your do-not-contact list is skipped — on calls, texts and campaigns.
Outbound calls skip do-not-contact numbers, hang up on answering machines and open by saying it's an AI, and outbound messages pass content safety checks before they go out.
Tight controls. Including on the AI.
Turn on two-factor and every sign-in asks for a one-time code sent to your email. Where you're signed in lists every device with its sign-in time, and you can sign any of them out. Turn on new sign-in alerts and you're emailed when your account is used on a new device. The activity log records who did what: you, your team, your API keys, and our support team if they ever open your account.
You choose what the AI may do on its own. Each action can run automatically, ask first (it waits in an approval queue) or never. Webhooks are signed with secrets you can rotate, and incoming Twilio requests are signature-verified.
Everyone signs in as themselves.
Invite your team as Admin, Staff or Viewer. Each person has their own sign-in and sees only what their role allows, checked by the server on every request. API keys and your own account stay with you, whatever the role, and only you can change another Admin.
Sign-in codes are set per person: new Staff and Viewers start with an emailed code at every sign-in, and you decide for each member. Remove someone and they lose access straight away, even on a page they have open, and you can sign any member out of their devices. Every change a member makes is in the activity log.
Security specifications
The details your security team wants.
Specific controls, all live in the product today.
Guardrails
The never list.
Guardrails built into the product, not promises on a slide.
Trust documents
Read the paperwork.
Public documents anyone can read. Have a security questionnaire? Email it to us.
Privacy Policy
What we collect, how we use it, and your rights as a data subject.
ReadData Processing Addendum
The terms that govern how we process personal data on your behalf.
ReadGDPR
How we handle GDPR data-subject rights and requests.
ReadSub-processors list
The third parties we use to run Aiva Chat, what they receive and where they're located.
ReadTerms of Service
The agreement that covers your use of Aiva Chat.
ReadSecurity questionnaire
Doing vendor diligence? Send us your questionnaire and we'll fill it in.
EmailDirect line
Talk to our security team.
Real humans. Real responses. For security questions, you reach us directly.
Email support@aivachat.io
For security questionnaires, vendor reviews, DPA questions, or anything else security-adjacent. Vulnerability reports (responsible disclosure) go to the same address — please include “[VULN]” in the subject.
Diligence done. Try it free.
Start with a 14-day free trial. No card. Questions first? Email our security contact.