Security
Alden is trusted with your computer and your logins.
Quant Intelligence
Report a security issue
Found a weakness in Alden? Tell us privately first, and give us a fair chance to fix it before anyone else hears about it.
What helps us most
- Your operating system and the Alden version
- The permission rules you had set at the time
- Whether an approval prompt appeared, and what it said
- For injected instructions, the page or message that carried them
- The relevant entry from the action log, which records what was actually tried
Email security@quantintelligence.co
How Alden keeps you safe
- Permission before actionEvery action goes through the Governance Engine before it runs, checked against the Always / Ask Every Time / Never rules you set for sign-ins, browsing and actions.
- Your machine, your dataThe model runs on your own device. Your files, your conversations and what you ask Alden to do stay on your machine rather than being sent anywhere to be processed.
- Bounded file accessAlden creates a folder called AldenWorkspace on your Desktop during setup. It reads what you put there and doesn't go looking elsewhere on your Mac. Skipped it during setup? Create it later in Settings → Permissions.
What Alden asks your Mac for
macOS makes an app ask before it can reach anything sensitive, and every prompt comes with a reason. This is the complete list.
- MicrophoneThe wake word and voice commands. Asked during setup.
- CameraHand gestures, which are optional. Asked during setup.
- AccessibilityControlling app windows for you. Asked during setup.
- AutomationDriving Mail, WhatsApp and Messages when you ask Alden to. Asked during setup.
- Desktop (Files & Folders)Creating the AldenWorkspace folder. Asked at the end of setup.
What it never asks for
- Full Disk AccessNothing Alden does needs it. To read your email it asks Mail, the way you would, rather than opening Mail’s database. It doesn't touch Messages’ store, Safari history or any other app’s private storage.
- Your photo libraryNot requested.
- Contacts, Calendar, Reminders and LocationNot requested.
You can decline any of them. The feature that needed it stops working, everything else carries on, and you can allow it later in Settings → Permissions.
The threat model
Alden runs on your own machine, holds the keys to your accounts and can act without you watching. That shapes what we defend against.
The model runs locally by default, so the usual cloud-assistant risk, your prompts and files sitting on someone else's servers, doesn't apply. The harder question replaces it: what can software on your machine be talked into doing? The Governance Engine exists for exactly that.
- FilesOnly what's in your Workspace folder, or what you grant. The rest of your drive is off-limits, so a prompt that tries to read elsewhere fails rather than succeeding quietly.
- SitesSigning in follows per-site rules you set. A site you never allowed isn't reachable just because a task mentions it.
- Actions that leave the conversationSending mail, moving files, purchases: each follows its own Always / Ask Every Time / Never rule.
- Never bypassedCAPTCHAs and legal agreements always come back to you, and payments wait for your spoken confirmation. Neither can be turned off by a task, a prompt or a forgotten setting.
- Prompt injectionPages and emails Alden reads can contain instructions. We treat that as a real attack: the actions that matter are gated whatever convinced Alden to try them, so injected text can't give itself file access or approve its own payment.
Your data
What's collected and how long it's kept. The developer API runs on our servers rather than your machine, and the policy covers it separately.