Question 1

What is Verillian?

Accepted Answer

A runtime AI governance layer for how your institution uses AI. A lightweight sentinel on each device governs what any AI tool or agent can do the moment it acts, keeps sensitive data inside your boundary, and writes a signed, tamper-evident record of every decision. Real-time control and the compliance evidence, end to end.

Question 2

Will it change how our people work?

Accepted Answer

No. People keep using the AI tools they already use. The only visible sign is a small status icon. There is nothing new to learn and no workflow to change.

Question 3

Which AI tools, agents, and providers does it cover?

Accepted Answer

Any AI tool or agent that reaches a provider over standard secure web traffic: chat assistants, coding assistants, autonomous agents, and command-line AI clients. Verillian detects, redacts, and blocks at the boundary across providers, with best-effort detection coverage. Audit entries are fully parsed for Anthropic today (Claude and Claude Code), and parsing expands to the providers our first institutions use.

Question 4

Is Verillian an AI company?

Accepted Answer

No. There is no model. Verillian is the layer the models pass through, where your policy is enforced and the record is made.

Question 5

What makes Verillian different from a VPN or cloud gateway?

Accepted Answer

A VPN or cloud gateway sits between your network and the outside world. It cannot see what is inside an HTTPS session, so it cannot decide what an AI tool may do or sign a record of what it did. Verillian runs on the device, where TLS terminates, and acts before anything reaches the provider.

Question 6

Can Verillian stop all AI traffic if needed?

Accepted Answer

Yes. The org-wide emergency stop halts all governed AI traffic across every enrolled device at once, on demand, with no per-device action required.

Question 7

Does Verillian read our content?

Accepted Answer

No. Audit content is encrypted under a key only you hold, on a server that stores ciphertext it cannot read. Verillian the company never receives your AI content.

Question 8

Is our data used to train AI models?

Accepted Answer

No.

Question 9

Once installed, can Verillian reach into our system?

Accepted Answer

No. It runs entirely in your environment and verifies its license offline, with no phone-home. We cannot see your activity, start it, or stop it. Commercial control comes from the signed license key, not from access.

Question 10

Where is the record kept, and who can see it?

Accepted Answer

In your environment, under your control. Only the administrators and reviewers you authorize can see the activity record.

Question 11

What devices are supported?

Accepted Answer

macOS and Windows, deployed through your existing device management tooling: Jamf, Intune, Kandji, Configuration Manager, or Group Policy.

Question 12

Do we need to change our AI tools or install plugins?

Accepted Answer

No. The sentinel works without changes to the tools themselves.

Question 13

Does it work in air-gapped or restricted environments?

Accepted Answer

Yes. Verillian operates, and verifies its license, without any dependence on outside connectivity.

Question 14

Will it slow things down?

Accepted Answer

No. The sentinel is lightweight and designed to stay out of the way.

Question 15

How long does deployment take?

Accepted Answer

Sentinels deploy through your existing endpoint management, like Jamf, Intune, or Group Policy. Enrollment is typically complete across a fleet in the same window as any other software push.

Question 16

What happens if a device loses connectivity?

Accepted Answer

The sentinel buffers events locally and uploads when connectivity returns. If the buffer fills, it fails closed: all AI traffic stops until the backlog clears. Ungoverned AI is never a degraded mode.

Question 17

If Verillian runs on our infrastructure, how is our device count enforced?

Accepted Answer

Through a signed license key. Verillian is self-hosted, so there is no remote control plane, but the software runs only what its key authorizes. The key encodes your device count and term, and the software verifies it offline on every start. Only Verillian holds the keys that sign it, so an altered or fabricated key fails verification.

Question 18

Can we add devices in the middle of a term?

Accepted Answer

Yes. Tell us the new count and we issue an updated key that runs to your existing renewal date. The added devices are prorated to the time left in your term, so you pay for the remaining months, not a full year. Reducing your count is handled at renewal.

Question 19

How is Verillian priced?

Accepted Answer

Per device, per year, not a one-time fee. Multi-year terms are available and are encoded directly into the license key. Contact us for pricing for your institution.

Question 20

How does renewal work, and what happens if we stop?

Accepted Answer

Before your term ends, you renew and we issue a new key. If you choose not to renew, no new key is issued and the software locks at expiry until a valid key is installed. Only Verillian can generate a valid key, so coverage is always explicit and under contract. Your data is untouched and stays in your environment throughout.

Question 21

How do we know the record was not altered?

Accepted Answer

Every entry is Ed25519-signed and hash-chained to the one before it, so any change to any entry is detectable. You can stand behind the record in an audit or review.

Question 22

Which compliance frameworks is it built for?

Accepted Answer

Regulated sectors including law enforcement, healthcare, education, government, defense, and financial services: CJIS, HIPAA, FERPA, CMMC, FedRAMP-aligned, and financial frameworks. Each maps to enforcement at the device and a signed record: CJIS Security Policy v6.0 tamper-evident logging and one-year retention, HIPAA data minimization and six-year documentation retention, and NIST 800-53 audit controls. Verillian is designed to satisfy these frameworks' controls and is aligned to their requirements, not independently certified. CJIS compliance is validated through your agency's state-agency audit. Control-by-control mappings are on the trust center.

Question 23

How do we report a security issue?

Accepted Answer

Email security@verillian.ai. See the vulnerability disclosure page on the trust center.

Question 24

Who can see the audit record?

Accepted Answer

Only administrators and reviewers your institution authorizes. The content is encrypted under your key; Verillian the company stores ciphertext it cannot read. Every decryption is itself logged, and you can grant time-boxed access to an auditor and then revoke it.

Question 25

What happens to our data if we stop using Verillian?

Accepted Answer

Your data stays in your environment. We do not hold it. When a license expires, no new key is issued and the software locks, but nothing in your environment is touched or deleted.

Questions, answered

About Verillian