FREE · NO SIGN-UP

Check a phishing email.
Before you click.

Is this email genuine or a scam? Paste the message and start the AI check with one click. Preparation takes place on your device; the configured AI then analyses the prepared version.

Check an email for phishing

Preparation takes place locally. Transmission to the AI starts only when you click “Check for phishing with AI”.

When you start, a locally prepared version of your email is sent to SkyMail and Google Cloud (Vertex AI) for analysis. A high risk score may result in a public warning page containing the cleaned subject, separately reviewed and cleaned email text, identified provider and Return-Path domain. Personal details are automatically removed for publication; email text without sufficient clearance remains hidden. Automated detection can make mistakes. You can preview the prepared version. Privacy policy

Text and HTML: up to 200,000 characters. Complete email source is detected automatically and may be up to 5 MB, with a maximum of 200,000 decoded text characters.

If you paste email text, add the subject here. It is extracted automatically from email source and EML files. It helps others recognise the same email; personal details are additionally removed for a public warning page.

View or edit the prepared version (optional)

Links, images and attachments are not opened. Certain personal details and URL paths are removed locally. Cleaning may be incomplete; you can edit the version before starting the AI. For a public warning page, the subject and email text undergo an additional review and cleaning for personal details. Published wording comes from this version; the AI does not rewrite it. Email text without sufficient clearance remains hidden. Privacy and storage

At a risk score of 70 out of 100 or above, a public warning page may be created automatically. It shows the cleaned subject and the separately reviewed, cleaned wording, without personal recipient details and with defanged links. If the text has not been sufficiently cleared for publication, only a content summary is shown. Your original file and private assessment report are not published.

Loading AI availability …

Local indicators alone do not provide an overall assessment. An AI assessment does not confirm authenticity or the sender’s identity either.

FROM SUSPICION TO ASSESSMENT

How to check a suspicious email

  1. Paste the email. Paste text, email source or an EML file. You can optionally preview and edit the prepared version locally.
  2. Start the AI check. Click “Check for phishing with AI”. Local preparation and the configured AI analysis then run automatically; links and attachments are not opened.
  3. Cross-check the assessment. Read the reasoning and uncertainties. Independently verify unexpected requests through the alleged sender’s known app or website.

Recognising phishing emails: typical warning signs

Common pretexts include a threatened account suspension, an outstanding parcel fee or an allegedly necessary confirmation of your login details. Even well-written messages with a personal greeting and a familiar logo can be forged.

The checker assesses such text patterns and examines embedded links statically. With pasted HTML or an EML file, it can detect a difference between visible link text and the actual destination. It does not visit the destinations.

Sender details and authentication headers in a submitted file may have been manipulated. The check confirms neither the sender’s identity nor SPF, DKIM or DMARC results.

Public phishing check results and message patterns · Learn more about phishing indicators · Advice from the German consumer advice centre

Frequently asked questions about the email phishing check

Is the phishing email check free?

Yes. You can check email text, links and supported EML files free of charge without signing in. You do not need a skymail mailbox.

Can the check tell for certain whether an email is genuine?

No. Local preparation reveals individual indicators. After your click, the configured AI can assess the prepared content but can also be wrong. A low warning score confirms neither the sender’s identity nor harmless links. If in doubt, ask through a contact method you already know.

Is this email genuine or an attempted scam?

Compare the request with what you expect: are you unexpectedly being asked to confirm details, sign in or pay? Check the message for warning signs in the tool, then open the alleged service independently through your known app or an address you enter yourself. An automated check alone cannot confirm authenticity.

How can I verify an email sender?

Display the complete sender address in your email client. The visible display name alone tells you little. For an EML file, the checker compares available From, Reply-To and Return-Path domains. Differences can be suspicious; matching addresses do not prove authenticity. Senders and supplied headers can be forged.

How can I check a link in an email for phishing?

Copy the actual link address without opening the destination and select “Link from an email” above. The checker examines the address statically for unusual features. Short links are not resolved and the destination is not visited. The check therefore cannot confirm that a link is safe.

Is an email from PayPal, DHL, Sparkasse, Amazon or Microsoft genuine or fake?

You can paste these messages too. The checker looks for suspicious text patterns and link destinations. It is independent of these providers and does not confirm messages on their behalf. Open your account or parcel tracking through your known app or the original address you enter yourself.

Can the AI also recognise previously unknown companies?

Yes. The AI can use the prepared sender name, subject and content to identify the company the email claims to represent. No previous catalogue entry is required. The name is reviewed separately before public use. It confirms neither the message’s authenticity nor its actual sender.

Is my email stored or sent to an AI?

When pasted, the message stays on your device. Clicking “Check for phishing with AI” requests analysis: the input is prepared locally and the version is sent to SkyMail and Google Cloud (Vertex AI), if AI is configured. You can optionally preview and edit it first. Cleaning does not guarantee complete anonymity. For high-risk messages, the cleaned subject and separately reviewed, cleaned email text may be stored on a public warning page. The wording is taken from the prepared message; personal details and individual URL components are removed. Shortened versions are labelled accordingly. Without sufficient clearance, the email text remains hidden. The original file, complete email headers and your private assessment report are not published. Processing and retention details are explained in the privacy notice.

What is transmitted when sharing?

Sharing an assigned warning page passes its public link and a general notice. The destination may contain the separately reviewed, cleaned email text. Message text, personal links and your private security report are not passed directly to the sharing function. The selected messaging or social media app processes sharing under its own rules.

What should I do if I already clicked?

Do not enter any more details or open downloaded files. If you entered login credentials, change them on the known original website from a trusted device. If payment details are involved, contact your bank immediately through a known contact method.

Do I need an internet connection?

You need an internet connection to load the phishing email checker and run the AI analysis. Input and preparation take place on your device. Transmission of the prepared version starts only when you click “Check for phishing with AI”.

Which messages are other people checking? See frequently reported phishing emails in the SkyMail checker.

SkyMail Phishing Email Check 2.13.0 · No guarantee of safety