Plain answers, not badge soup.
All traffic — website, dashboard, API and snippet — is encrypted with TLS. No exceptions.
Billing runs entirely on Stripe's certified infrastructure (PCI DSS Level 1). We never see or store card numbers.
Screened lead data is automatically deleted after 90 days by default. Agency plans set their own window in the dashboard — any number of days, down to zero, where lead details are cleared at the next nightly sweep.
Keys can be regenerated instantly from your dashboard; old keys die immediately. We alert you by email on every key change.
Device fingerprints are hashed in the visitor's browser before they reach us. No third-party tracking CDNs are loaded on your site.
If LeadShield is unreachable your forms submit normally. Our availability can never block your business.
Form fields are an open door into the AI assistants businesses now attach to their CRMs. On every plan, LeadShield flags known prompt-injection patterns (in several languages, including text hidden in invisible characters) and risky links, and on Pro and Agency your webhook carries a cleaned copy of each message for any AI step. It flags; it can't promise to stop every attempt, so keep a person in the loop wherever an AI tool can act on a lead.
UptimeRobot checks LeadShield's API around the clock and publishes the result, with its uptime history, on our public status page: stats.uptimerobot.com/fK5aATzZdh. If something is wrong, it shows there.
These companies process data for LeadShield, each only for the job listed:
| Company | What for | What it receives |
|---|---|---|
| Hetzner | The server LeadShield runs on | Everything LeadShield stores |
| Hostinger | Off-site backups | Encrypted copies only, which it can't open |
| Cloudflare | Network, DNS and bot checks on our own sign-up and contact forms | Traffic to leadshield.ai |
| Google (Gemini API) | AI analysis of lead content, on Pro, Agency and in your first 14 days, the Lead Brief, the free spam message checker, and the site's chat assistant | The lead's name, company, email, phone and message, for that one check, plus your Business Profile for a Brief; text pasted into the spam message checker; what you type into the chat |
| Google (Sign in with Google) | Signing in, only if you choose it | Nothing from us. Google tells us your name and email address |
| AbuseIPDB | IP reputation | The visitor's IP address only |
| Brevo | Account and service emails | Your email address and the email's contents |
| Stripe | Payments | Billing details; we never see card numbers |
| Google Analytics | Statistics about visits to leadshield.ai | Only for visitors who accept cookies |
On 6 October 2026 an automated attack used the sign-up form on leadshield.ai to create 257,087 fake accounts in 23 hours. Each sign-up sent a verification email, so our domain mailed those emails to addresses the attacker chose, with a link typed into the name field. In the same incident our public chat assistant was used as a free AI model, and one of our AI keys was used outside LeadShield.
What we changed:
We've left out anything that would help someone repeat it. Questions about it are welcome through the contact page.
LeadShield.ai is built and operated by Autonoma, 66 Paul Street, London, EC2A 4NA. Card statements show AUTONOMA.
Security question or something to report? Contact us — security reports get same-day attention.