Link safety

How to check an unfamiliar shared link

Do not let urgency choose for you. Verify the message, reveal the destination and decide whether you need to open it at all.

Stop if the link asks for credentials, payment, software installation or disabled security controls. Navigate to the known service independently instead.

Begin outside the URL

Ask whether you expected the message and whether the sender normally uses this channel. A familiar display name is not proof: email addresses can be spoofed, accounts can be compromised and chat profiles can be copied. Verify unusual requests through a phone number or conversation you already trust—not through details in the suspicious message.

Urgency, secrecy and a threat of account closure are signals to slow down. CISA’s phishing guidance recommends recognising common signs and reporting suspicious messages rather than interacting with them.

Inspect the destination

On a desktop, hover without clicking and read the status-bar URL. On a phone, a long press may reveal a preview, but avoid opening an unsafe preview automatically. For a shortened link, use the provider’s documented preview feature if one exists or ask the sender for the full destination.

Read the registered host from right to left. In accounts.example.com.evil.test, the controlling domain is evil.test, not example.com. Look for subtle substitutions, extra words and international characters that resemble Latin letters. HTTPS only means the connection to that host is encrypted; it does not make the host honest.

Separate viewing from downloading

A direct download deserves more caution than an ordinary page. Unexpected archives, disk images, scripts, macro-enabled documents and installers should not be opened on a normal workstation. Confirm the file through a trusted source and use organisation-approved scanning or an isolated analysis environment. Do not upload confidential documents to a public scanning service.

Browser and operating-system warnings are not obstacles to work around. If a message tells you to disable protection, paste a command into a terminal or install a “viewer,” stop and ask your IT or security contact.

Use an independent route

If the message claims there is a problem with a bank, cloud drive or social account, open the official app or type the known address yourself. Check notifications from inside the service. This avoids both the suspicious link and a convincing imitation page.

For public web addresses, the URLhaus project publishes information about malware-distribution URLs. A clean lookup is not a guarantee—new attacks appear before blocklists update—but a positive match is a strong reason to stay away and report the message.

Report without spreading

Use the reporting tools in your email or messaging service. In a workplace, forward the original message according to policy so headers are preserved. Avoid reposting a clickable malicious URL in a public warning; use a screenshot or defanged form if your security team asks for one.

The safest click is often no click. Most legitimate senders can explain what they sent, provide the full destination and wait while you verify it.