---
title: "Troubleshooting & FAQ"
description: "Read the Sonomos Support Center guide for Troubleshooting & FAQ. Nothing's showing up, the engine won't start, and the other usual suspects."
url: "https://sonomos.ai/support/locke/troubleshooting/"
source: "https://sonomos.ai/support/locke/troubleshooting/"
---

Reference

# Troubleshooting & FAQ

The problems people actually hit, with the checks that resolve them — ordered by how often each turns out to be the cause.

If you're troubleshooting before you've read anything else, the basics assumed below — what Locke watches, what masking means — are covered in [What Locke is](https://sonomos.ai/support/locke/what-it-is/).

## Nothing is showing up in the feed

By far the most common report, and it's usually not a fault. Work through these in order:

1. **Is protection on?** The Dashboard status card should be green and read Protected. Grey means stopped; amber means paused, with a countdown in the titlebar.
2. **Is the app you tested actually protected?** Check Controls › Protected surfaces. If it's in the Detected, not yet managed card at the top, it isn't covered yet — press + Protect.
3. **Was there anything to catch?** Send something unambiguous — your own email address is ideal. A message with no recognizable sensitive data correctly produces no detections.
4. **Is the relevant data type switched on?** Check Controls › Detection data types. The five region-specific types ship off — see [What Locke detects](https://sonomos.ai/support/locke/data-types/).
5. **Prove the detector works at all.** Open the [Sandbox](https://sonomos.ai/support/locke/sandbox/) and press Run test on the pre-filled sample. If that finds nothing, the problem is the engine, not your configuration — see below.

Tip

The Sandbox is the fastest way to split "Locke isn't detecting this" from "Locke isn't watching that app". If the Sandbox catches it, detection works and the issue is coverage. If it doesn't, the issue is the detector or your data-type settings.

## Protection won't start

| Symptom | Likely cause | Fix |
| --- | --- | --- |
| Start protection does nothing, or flips back to stopped | The detection engine isn't installed or can't be found. | Re-run setup from Support › Replay the welcome guide, which re-checks for the engine and installs it if needed. |
| Detector unavailable in the Sandbox | The engine isn't running. Live protection isn't working either. | Restart the app. If it persists, reinstall from sonomos.ai/downloads — your settings and history survive a reinstall. |
| It worked yesterday and not today | An OS update, or security software quarantining the engine. | Check whether your endpoint protection has flagged Locke, then reinstall. |
| It starts and then stops on its own | A crash loop. The app tries to restart it a few times before giving up. | File this one — [Support](https://sonomos.ai/support/locke/support/), with diagnostic logs attached. It's not something to work around. |

## Websites aren't protected, but desktop apps are

Almost always the certificate. To read what your browser sends before it's encrypted, Locke installs a local certificate during setup. If you declined it, or it was removed, browser traffic can't be inspected — while desktop AI apps keep working normally.

Re-run setup from Support › Replay the welcome guide and accept the certificate step. It's generated on your machine and never leaves it.

## Locke masked something it shouldn't have

1. Find the item in Reports › Detection & masking history and note its data type.
2. If the type is one you don't need — an Organization Name in a context where company names are the point — switch that type off in Controls › Detection data types.
3. If the type is right but the match was wrong, lower Controls › Scanning floor from Strict to Balanced, or from Balanced to Relaxed.
4. If it was one of your own custom rules, the pattern is too broad. Test a narrower version in the [Sandbox](https://sonomos.ai/support/locke/sandbox/) before saving it.

Prefer narrowing to switching off

Turning a whole data type off is a big hammer — it means that category goes out in the clear from then on. Lowering the sensitivity floor, or fixing one over-broad custom rule, usually solves the actual problem without opening a hole you'll forget about.

## The numbers look wrong

| Symptom | Explanation |
| --- | --- |
| The count seems far too low | By default only You typed is counted. Everything your AI tool attaches on your behalf is protected but not tallied. Turn more sources on in Settings › What counts as a detection. |
| The count jumped enormously | You enabled an extra source — most likely Injected context, which is usually the largest by a wide margin. Nothing about your protection changed. |
| A report disagrees with the Dashboard | It shouldn't — both honour the same counting settings. If they differ, the settings changed between generating the two. Regenerate the report. |
| Items caught and items hidden differ | Something was detected but not masked — usually caught while paused, or a failed substitution. The history shows which. |
| History is shorter than expected | If you set a retention limit in Settings, anything older has aged out. Export before you need it. |

## A setting is greyed out

You're on a Teams plan and your admin locked it. The control names who set it. Only they can change it — Locke support can't override org policy. See [For team members](https://sonomos.ai/support/locke/members/).

## Frequently asked

### Does Locke slow anything down?

Detection adds a small delay before a message is sent, on the order of milliseconds for text. Files and voice cost more, which is why they ship switched off — turn them on if you need them.

### Does it work offline?

Detection and masking run on your machine and don't send your content anywhere, so inspecting a message needs no network round trip. The app does need to be online to sign in to your Sonomos account and to periodically confirm your licence (and, on Teams, your seat and policy updates). In practice you're online anyway — the AI tools Locke screens need the internet.

### Does closing the window turn protection off?

No. Locke keeps running in the background with the corner dot visible. To actually stop it, use the pause menu — see [The window](https://sonomos.ai/support/locke/tour/).

### Does signing out turn protection off?

No. Locke protects the device, not the login. Signing out disconnects your account — a Teams member stops receiving policy — and local detection carries on with the settings already on the machine.

### Will the AI get confused by the fake values?

Rarely, because the stand-ins keep the format and are stable — the same input always produces the same substitute, so an AI can follow references across a conversation. Where it does matter is a task about the value itself: "is this card number valid" can't be answered about a substitute. For those, use the [Sandbox](https://sonomos.ai/support/locke/sandbox/) to see what's being swapped, then decide whether to narrow a data type for the duration.

### Can I get the original values back?

Your originals never left your machine — you already have them; only the copy sent to the AI was substituted. Locke doesn't rewrite what you typed in the AI tool's own interface.

### Can Sonomos see what Locke catches?

No. Detection is local, and the values Locke finds are never uploaded. Your history is a local log storing hashes rather than plaintext. If you attach diagnostic logs to a support request, those contain operational detail and counts — never the content of what you sent.

### Can my admin read my prompts?

No. Admins see counts and policy compliance. The full answer, with the reason it can't change, is at the top of [For team members](https://sonomos.ai/support/locke/members/).

### Can I use Locke on more than one computer?

Yes — install it on each. Settings and history are per-device, so numbers and reports are per-machine too. A Teams seat covers the person, and their policy applies to every device they sign in on.

### Is Free really free?

Yes. It's the complete protection engine for one person, not a trial. The one limit is reporting: one basic Privacy summary a month. Teams adds full reports and org management, not better protection. See [Free vs Teams](https://sonomos.ai/support/locke/plans/).

Security & technical documentation

Security, compliance readiness, sub-processors and technical architecture are all documented on the Sonomos Trust Centre: [trust.sonomos.ai](https://trust.sonomos.ai) (access-restricted during beta).

## Still stuck

Go to [Support & feedback](https://sonomos.ai/support/locke/support/), leave Attach diagnostic logs on, and include your OS, the app version from Settings › Diagnostics & updates, and what you expected to happen. Don't paste the sensitive value itself — describing the kind of value is always enough.