Exchange anti-phishing codes: setup and safe checks
Last reviewed:
An anti-phishing code is a personal recognition string configured inside an exchange account. The exchange includes it in the communications covered by that feature, helping you spot messages that do not match your settings. It is not a login code, an authenticator seed or permission to trust a payment request. Use it as a warning signal alongside independent navigation to the official service.
What an anti-phishing code can and cannot establish
| Check | Useful signal | Remaining limitation |
|---|---|---|
| Correct code | Matches a string configured in your account | Can be copied from exposed messages |
| Missing or wrong code | Warning for a channel that should include it | Coverage and account context still need checking |
| Email sender and layout | Additional context for investigation | Names, logos and links can be imitated |
| Independent app check | Lets you review the real account directly | Requires a trusted app, device and login |
Step by step
Open the verified account
Use the official app or a trusted bookmark. Do not begin from a message that pressures you to reset security or save funds.
Read the coverage description
Find anti-phishing protection in security settings and check whether it covers email, SMS or specified message categories for your account.
Create a separate recognition string
Follow the displayed format rules. Do not reuse a password, identity number, recovery phrase or authenticator seed as your code.
Confirm and inspect safely
Finish the account verification and check the saved status. Observe a relevant notification from an ordinary action you initiated, without making an unnecessary transfer.
Act through independent navigation
If a future message is unexpected, missing the code or showing a wrong code, open the official service separately. Report suspected phishing through its verified support channel.
What the code adds to message checking
A familiar sender name, logo and polished email layout are easy to imitate. A recognition string you configured gives you another detail to compare. If a message that should contain the code does not show it, stop before following its instructions. A wrong code deserves the same caution. The next action is to inspect the real account through a trusted entry point, not to reply to the suspicious sender and ask for reassurance.
The feature is useful precisely because checking it is simple, but simplicity has a boundary. The string is displayed in messages and can become visible to anyone with access to those messages. Treat it as private recognition information rather than a secret that can safely authorize transactions. A copied code can appear in an attacker’s message without making the message authentic.
Email and SMS coverage are service-specific
Do not assume every exchange inserts its anti-phishing code into every channel. Binance’s official anti-phishing guide describes both email and SMS coverage. OKX’s account security guidance describes the code as an aid for recognizing email. Other services may apply it to specified email categories or offer different behavior across products. Read the official description for your account and channel instead of transferring one exchange’s claim to another.
A text message appearing in the same conversation as older genuine messages is not conclusive proof of origin. Sender identifiers and conversation grouping can mislead. If SMS coverage is not confirmed for your service, the absence of a code cannot by itself classify every text. Independently verify unexpected messages anyway, especially those asking you to change security settings, install software or move funds.
Choose a recognizable string without exposing other secrets
Choose something you can reliably recognize that is not your exchange password, email password, authenticator seed, wallet recovery phrase or identity document number. Follow the character and length rules shown in the real setup screen. This guide intentionally provides no universal format because platforms can use different constraints and update them. A password manager can hold a private record of which recognition string belongs to which exchange.
Separate codes across services make it easier to notice an incorrect context and reduce reuse after one mailbox leak. Do not publish a screenshot containing the code, put it in a public support thread or include it in a social media profile. If a screenshot is needed for a support case, remove the code and other sensitive fields first. Support does not need your unrelated login credentials to understand that a message looks suspicious.
Set up through the authenticated account
Open the official app or a verified bookmark and enter the account security area. Find the anti-phishing feature, read which messages it covers and create the recognition string there. Complete whatever security verification the account requests. Do not open the setting through an unsolicited email saying that your code has expired. The security change should be an action you start deliberately inside the real account.
After saving, check that the setting shows the expected value or enabled state. Use the next relevant message generated by an ordinary action you initiated to understand how the code is displayed. Do not perform an unnecessary withdrawal simply to generate a sample. A normal account notification may be enough if it belongs to a covered category. Keep the distinction between verifying delivery behavior and proving the safety of all future communications.
Where to begin on Bitget, OKX and Binance
For Bitget, start from the verified profile or personal account area, open security and look for the anti-phishing setting. For OKX, enter account security and look for anti-phishing protection, then read the email coverage explanation. For Binance, use the account security area and locate the anti-phishing code setting. Follow the official page (以官方页面为准). These generic paths avoid depending on an old screenshot or a menu label that differs on mobile.
Before changing a code, check whether you are looking at the correct account and regional service. An email from an unrelated product may not use the same recognition feature. A search result with a familiar brand does not establish that you are in the right account. Reach the help center from the official service you independently opened, and compare its instructions with the security controls actually displayed.
Where to begin on Bybit, MEXC and Gate
For Bybit, use the verified account security area and search the official help center for anti-phishing code setup. For MEXC, open account security and use the official anti-phishing code help entry if the control is unclear. For Gate, look in the account security area and consult its security guide for anti-phishing setup. Follow the official page (以官方页面为准). Availability, accepted strings and confirmation challenges must be checked in the current account.
If you cannot find a setting, do not install an extension or scan a stranger’s QR code to unlock it. Ask support through the already verified application whether the feature is available for your account. The absence of an optional recognition code is a reason to be more careful with messages, not a reason to replace official navigation with a third-party setup service.
Use a message inspection routine
First ask whether you expected the message. Then check the code where the service says it should appear, the sender details and the requested action. A claim that your assets will be frozen unless you act immediately deserves independent verification. Open the app separately and inspect account notifications, security events or support messages. Do not treat a good-looking address in the email body as proof that the actual link goes there.
Even a correct code cannot validate an attachment, a telephone number or the destination of a transfer. It also cannot establish that an investment offer is suitable or that a support representative is genuine. If the message asks for a wallet recovery phrase, remote access or a transfer to a verification wallet, stop. The recognition string does not transform those requests into normal account maintenance.
Handle missing, wrong or leaked codes
A missing code in a covered message is a warning, not a diagnosis of exactly what happened. The message may be fraudulent, the account may use another configured string, or the message category may fall outside the feature. Do not resolve uncertainty by clicking the message’s own recovery link. Compare the setting inside the real account and ask official support if the discrepancy remains.
If someone else saw your code, change it through the authenticated security page. A mailbox compromise requires more than changing the recognition string: secure the mailbox, inspect forwarding rules and sessions, review exchange access and check for unauthorized security changes. Preserve suspicious message details for an official report, while keeping codes and personal data out of public posts. A changed string does not invalidate a stolen account session.
Use the feature within a wider security routine
The code helps with message recognition. An authenticator or passkey protects authentication, while an address allowlist can restrict withdrawal destinations. These controls address different points in an attack. Maintain an updated device, use unique passwords and protect the recovery email. When changing phones or email accounts, verify that your recognition routine still works and that old account access has been removed appropriately.
Avoid training yourself to act automatically whenever the code matches. The strongest habit is to use messages as notifications and the independently opened official service as the place to act. That habit remains useful if the code is copied, a sender is spoofed or a service changes its communication format. An effective check should slow down an unexpected request enough for you to understand what you are authorizing.
Frequently asked questions
Is the anti-phishing code the same as a 2FA code?
No. It is a recognition string shown in covered communications. A one-time authenticator or SMS code verifies an action. Never substitute one for the other or share a live verification code with a person.
Does every exchange put the code in SMS?
No universal rule applies. Binance documents email and SMS coverage, while other guides may describe email only. Check the official feature description for the exact service and account.
Can I click a link when the code matches?
A matching code does not prove the link is safe. The code may have been copied. Open the official app or a verified bookmark independently before taking a sensitive action.
Should I change the code after a mailbox leak?
Yes, once the mailbox and account access are secured. Also review sessions, forwarding rules and security changes. Changing only the recognition string does not fix a compromised mailbox or exchange session.