Question 1

Will it hide a real email?

Accepted Answer

No — quarantined mail goes to a Spam folder with a score and a plain-English reason, and one tap marks it Not-spam (and trusts that sender for good). Nothing is silently deleted.

Question 2

Does scoring happen on my device?

Accepted Answer

No. Because email is untrusted input, scoring runs server-side at sync where a malicious message can't manipulate it. Only genuinely borderline messages get an AI second opinion.

Question 3

Standard vs Strict sensitivity?

Accepted Answer

Standard keeps borderline senders in your inbox but flagged; Strict quarantines them. Switch anytime in Settings → Rules → Spam.

Question 4

Can AI auto-sort my inbox?

Accepted Answer

Yes — every clean message is auto-categorised the moment it lands, server-side at sync. Mail is sorted into Primary, Social, Promotions, and Updates, so your unified inbox shows only what's actually personal while the rest waits in its own tab. The sorting reads sender, content, and context rather than crude keyword rules, so newsletters, notifications, and social pings file themselves. You can correct any call in a tap, and the system respects your allow-lists and rules on top.

Question 5

How does AI detect phishing emails?

Accepted Answer

AI Emaily scores phishing deterministically, server-side, before a message reaches your inbox view. It parses the authentication chain (SPF, DKIM, DMARC) to catch spoofed senders, weighs brand-impersonation signals, urgent-money and scam language, deceptive or mismatched links, and risky attachments, plus sender reputation. Only genuinely borderline cases get an AI second opinion — the model is a tie-breaker, never the sole judge. Anything that crosses the line is quarantined with a score and a plain-English reason you can read and reverse.

Question 6

Is this better than Gmail or Outlook spam filters?

Accepted Answer

It's a complementary layer that catches what they miss. Provider filters are strong on volume spam but routinely let through targeted phishing and lookalike-domain impersonation. AI Emaily re-scores every message with explicit auth-chain validation and impersonation detection, flagging spoofed and deceptive senders the provider waved through. And unlike a provider's black box, every decision shows a score and a reason, so you're never guessing why something was filtered — or losing a real email to silence.

Question 7

Will it accidentally block important emails?

Accepted Answer

It's built specifically not to. Nothing is ever silently deleted — flagged mail goes to a Spam folder with a score and a plain-English reason, and one tap marks it Not-spam, which rescues the message and trusts that sender for good. On Standard sensitivity, borderline senders even stay in your inbox, just flagged. Allow-lists let you protect senders permanently in advance. The design favours transparency and recovery over silent, unrecoverable filtering.

Question 8

Where does the spam scoring happen?

Accepted Answer

Server-side, at sync — never in your browser. Because email is untrusted input, scoring must run somewhere a malicious message can't manipulate the judgement or trigger code in your client. So every message is scored before a single byte reaches your inbox view. The deterministic scorer does the bulk of the work; only borderline messages (roughly 0.5–0.8) spend a metered AI token for a second opinion, keeping it fast and cheap.

Question 9

What are Primary, Social, Promotions, and Updates?

Accepted Answer

They're the four categories your clean mail is auto-sorted into. Primary is genuinely personal mail — the only thing your unified inbox shows by default. Social collects notifications from networks and communities. Promotions gathers marketing and deals. Updates holds receipts, confirmations, and automated notices. Each fills its own tab, so the inbox stays focused on what needs a human while everything else is one click away when you want it. You can recategorise anything in a tap.

Question 10

Why was an email marked as spam?

Accepted Answer

Because it's never a black box — open any quarantined message and you'll see a score plus a human-readable reason: failed authentication, a lookalike sender domain, scam or urgent-money language, a deceptive link, a risky attachment, or poor sender reputation. If the call was wrong, tap Not-spam to rescue it and trust that sender permanently. You can also tune sensitivity or add allow/block rules in Settings → Rules → Spam so it doesn't happen again.

Question 11

Can I whitelist or block specific senders?

Accepted Answer

Yes — the edges are always yours. Allow-list a sender or domain to ensure their mail always reaches you (marking something Not-spam does this automatically and permanently), or hard-block a domain to quarantine it outright. These lists sit on top of the automatic scoring, so you stay in charge of the cases the system shouldn't decide for you. Manage them all in Settings → Rules → Spam alongside the sensitivity setting.

Question 12

Does spam protection cost extra?

Accepted Answer

Core spam and phishing protection with auto-categorisation is part of the experience across plans, including the free $0 tier — keeping your inbox safe shouldn't be gated. The AI tie-break on borderline messages is metered against your plan, but clean and clearly-bad mail never spends a token because the deterministic scorer handles them. Pro at $17.99 and Autopilot at $29.99 unlock the wider agent; the protective baseline is there from the start.

Question 13

How does it handle malicious links and attachments?

Accepted Answer

They're scored as part of the deterministic pass before the message reaches you. Deceptive links — mismatched display text, lookalike domains, redirect chains — and risky attachment types raise the threat score and can push a message into quarantine with the reason spelled out. Beyond scoring, AI Emaily blocks tracking pixels and sandboxes links so you're not silently surveilled or one careless click from harm, and AI output is validated before render so nothing injects raw HTML.

Question 14

Will tracking pixels still load in my email?

Accepted Answer

No. AI Emaily blocks tracking pixels by default, so senders can't silently confirm you opened a message or harvest your IP and device details on open. Links are sandboxed rather than fired blindly. This sits alongside the spam and phishing scoring as part of treating email as untrusted input — the goal is an inbox that's not just sorted and safe from scams, but private by default.

Question 15

How quickly does it adapt to new spam?

Accepted Answer

Immediately, because the heavy lifting is deterministic. Auth-chain validation, impersonation signals, scam-language patterns, and link analysis don't need retraining to catch a fresh wave — and the AI tie-breaker reasons about borderline messages on the fly rather than relying on a stale blocklist. If you're under a targeted phishing push, flip to Strict sensitivity to quarantine borderline senders, and your Not-spam and block decisions take effect from the next message on.

A pre-triaged inbox, before you open the app.

Judged where it can't be tricked.

Protection that explains itself.

Never silently locked out.

From arrival to sorted, in one pass.

A calmer, safer inbox by default.

Where the shield earns its keep.

Good to know.

Let the noise sort itself out.