MetaMask Address Poisoning Protection Guide

Address poisoning exploits a routine many crypto users barely notice: copying a recipient from transaction history and checking only a few visible characters. An attacker creates a lookalike address, places it in the victim’s history with a small or zero-value transfer, and waits for that address to be copied during a later payment. MetaMask’s recipient checks and security alerts can identify suspicious patterns, but they do not replace verification of the complete destination.

Last reviewed: October 4, 2026. MetaMask interfaces, supported networks, alert providers and classifications can change. Confirm the current behavior in the official wallet and Help Center. This independent guide is educational, and no warning system can guarantee that every malicious address will be detected.

Essential rule: never choose a recipient solely because its first and last characters resemble a familiar address. Obtain the destination from an independent trusted source, compare the full address, confirm the network and use a test transfer when the amount justifies it.

What is a MetaMask address poisoning attack?

Address poisoning is a social-engineering attack against transaction-history habits. It does not normally require the attacker to break MetaMask, discover a private key or alter the blockchain. Instead, the attacker tries to make a fraudulent address appear familiar.

Blockchain addresses are long hexadecimal or encoded strings. Wallet interfaces often shorten them to a form such as 0x12ab…89cd. An attacker can generate many addresses until one shares a visually convincing prefix and suffix with an address the target previously used. The middle remains different, but that difference may be hidden in a shortened display.

The attacker then sends a tiny amount, a zero-value token transfer or another onchain interaction involving the target. This entry places the lookalike address in an explorer or wallet history. If the victim later copies the address from that poisoned entry, funds go to the attacker. A blockchain transfer to a valid but unintended address generally cannot be cancelled or reversed.

How the attack works step by step

Observation: the attacker monitors public blockchain activity and identifies an address that regularly sends assets to another wallet, exchange deposit address or treasury.
Lookalike generation: software generates addresses until one resembles the legitimate recipient at the beginning and end.
History poisoning: the attacker creates a transaction that causes the fraudulent address to appear in the victim’s activity history.
Visual shortcut: the wallet or explorer abbreviates both addresses, making them appear similar at a glance.
Copying error: during a later payment, the victim copies the poisoned address from history instead of returning to a verified source.
Irreversible transfer: the network executes exactly the address that was signed, even though it was not the destination the sender intended.

The malicious entry can arrive days or weeks before the attempted theft. Its purpose is not necessarily to steal immediately; it is to become the most recent or most recognizable entry when the user later searches transaction history.

Address poisoning is not the same as clipboard malware

Both threats can redirect a transfer, but the mechanism differs:

Address poisoning places a deceptive address in a public activity trail and relies on the user copying the wrong entry.
Clipboard malware runs on the device and replaces an address after the user copies it.
Phishing presents a fake website, support conversation, QR code or payment instruction containing the attacker’s address.
Account compromise may let an attacker change saved withdrawal details or control the communication channel used to provide an address.

The final defense is similar in every case: compare the address displayed in MetaMask immediately before signing with a destination obtained through a separate trusted channel. Verifying only the clipboard before opening the confirmation is not enough, because replacement can occur between steps.

How MetaMask address poisoning protection works

MetaMask announced address-poisoning detection in its send flow in June 2026. The protection is designed to notice suspicious recipient relationships while the user prepares a transfer. MetaMask’s broader security system also uses onchain analysis, ecosystem intelligence and security partners, including Blockaid, to classify risky addresses, tokens, websites and transactions.

A warning is a decision aid. Depending on the wallet version, network, account and available intelligence, MetaMask may display a recipient warning or additional context. The user must pause, obtain the address again from a trusted source and compare it carefully. The absence of an alert does not prove that an address is correct.

Why detection cannot be perfect

A newly generated attack address may have little or no prior reputation. A legitimate new recipient can also lack history. Threat intelligence changes as providers discover campaigns, receive reports and correlate onchain behavior. Detection therefore involves signals and confidence, not mathematical proof of the sender’s intent.

MetaMask cannot know which address a user meant to paste. If a fraudulent address has never been reported and does not match a recognizable poisoning pattern, the wallet may process it as an ordinary valid destination. Manual verification remains necessary even when all security features are enabled.

What MetaMask security labels mean

Label or state	Practical meaning	Recommended response
Verified	MetaMask has high confidence that the token, address or website matches a recognized official identity.	Still confirm the intended recipient, network and amount. Verification is not an endorsement or value guarantee.
Warning	Risk indicators or insufficient confidence require additional review. An address may be new, unverified or associated with suspicious behavior.	Stop. Re-obtain the destination independently and investigate every discrepancy before continuing.
Malicious	Strong threat signals associate the address, token, website or transaction with harmful activity.	Do not send, sign or connect. Close the site and verify the incident through official support channels.
No alert	No warning was shown for this request with the intelligence available at that moment.	Do not interpret silence as approval. Complete the same full verification.

A verified badge indicates identity confidence for the item being labeled. It does not certify an investment, promise that a contract is bug-free or confirm that this is the recipient you personally intended.

How to verify a MetaMask recipient safely

Return to the authoritative source. Open the exchange deposit page, hardware-wallet receive screen, verified project interface or direct message from a known recipient. Do not begin with a transaction-history entry.
Confirm the network. Ethereum, an EVM network and a non-EVM chain can use different address formats and deposit requirements. The asset name alone does not establish the correct network.
Compare the complete address. Expand or paste both values into a trusted offline text comparison where appropriate. Check every character, not only four at each end.
Check the MetaMask confirmation. Compare the destination shown in the final wallet prompt with the trusted source again. This catches clipboard replacement and last-step changes.
Review amount and asset. Confirm the number of tokens, decimals and fiat approximation. A correct recipient does not make an incorrect amount safe.
Review alerts without rushing. Expand warning details. Do not dismiss an alert merely because a website or caller says it is normal.
Use a test transfer when appropriate. Send a small amount, wait for confirmation and verify receipt through the destination—not merely through a screenshot.
Recheck before the main transfer. Use the address that the receiving service confirms, and repeat the final comparison. Do not assume the test makes a separately pasted main transaction safe.

Why checking only the first and last characters fails

Lookalike generation is specifically optimized for the characters people inspect. Matching more characters requires more computation, but automated generation can search large numbers of candidate addresses. A convincing prefix and suffix therefore provide no proof that the hidden middle is correct.

Checksum capitalization in an Ethereum address can detect some typing errors, but it does not establish ownership or intent. An attacker can generate a valid checksummed address. Treat a checksum as an error-detection feature, not an identity system.

When a test transfer helps—and when it does not

A test transfer can confirm that a selected address and network reach the expected destination. It is especially useful for a new exchange deposit, a large treasury movement or an unfamiliar cross-platform route. However, it only helps if the receiver independently confirms arrival and the main transfer uses the exact verified destination.

A test does not protect the next transaction if the sender returns to poisoned history and copies a different address. It may also involve minimum-deposit rules, memos, tags, bridge contracts or changing deposit addresses. Read the receiving platform’s current instructions before testing.

Safe workflows for common transfer types

Sending to your own second wallet

Display the receive address on the destination wallet itself. For a hardware wallet, verify the receive address on the trusted device screen when that functionality is available. Compare it with the address in MetaMask, confirm the chain, send a small test and verify the resulting balance independently.

Depositing to a centralized exchange

Open the exchange through a trusted bookmark or official app. Select the exact asset and network, then copy the current deposit address from that screen. Check whether a memo or destination tag is required. Do not reuse an old address from MetaMask history without verifying that the exchange still assigns it to your account.

Paying another person or business

Confirm the address over a second authenticated channel, particularly for a first payment or changed instructions. If an invoice arrives by email, verify it through a known phone number or established account rather than replying to the same possibly compromised message. Ask the recipient to confirm the test receipt.

Sending to a smart contract, bridge or dapp

Do not manually copy a contract from wallet history. Begin at the project’s verified domain and confirm contract addresses against official documentation. Review whether MetaMask is showing a simple transfer, contract interaction, token approval or signature. A familiar domain appearance does not make an unexpected permission safe.

Dust and zero-value transactions

A tiny incoming transfer does not mean the wallet itself has been hacked. Public addresses can receive unsolicited tokens and transactions without permission. The danger arises when the entry is trusted, copied or used to open an unverified link.

Do not send the asset back to an address merely to “remove” it. Do not visit a URL embedded in a token name or NFT description. Hide spam assets if the interface permits, and rely on verified sources for future recipients. The blockchain record itself cannot normally be deleted from public history.

Security Alerts, Basic Functionality and privacy

MetaMask’s current Help Center says security alerts are enabled by default in Extension and Mobile. For accounts onboarded with a Secret Recovery Phrase, the relevant controls are associated with Basic Functionality in privacy settings. Turning off that broader setting can also disable transaction simulations, phishing detection and other wallet features. Accounts created through some social login methods may keep the feature enabled.

For security checks, MetaMask states that relevant request information can be sent to a MetaMask server for EVM activity or to Blockaid for non-EVM activity. This can include the dapp URL, JSON-RPC method, chain ID and transaction data. MetaMask says Secret Recovery Phrases, private keys and credentials are not shared with security providers.

This creates a practical tradeoff: threat checks require request data, while a pared-back configuration reduces connected services. Users should review the current MetaMask privacy explanation rather than assuming either that all wallet activity is private or that security providers receive secret keys.

Security alert versus transaction simulation versus Added Protection

Layer	What it addresses	What it does not prove
Recipient/address alert	Signals that the destination may be suspicious, malicious or part of a poisoning pattern.	It cannot know every new attacker or guarantee that an unflagged recipient is intended.
Transaction simulation	Estimates asset and permission changes that a proposed action may produce.	A preview alone historically did not force actual execution to match, and simulations have limits.
Added Protection	On supported EVM smart accounts, enforces the protected preview outcome or causes the transaction to revert.	It does not correct an address the user intentionally approved or eliminate possible gas cost on a revert.
Token-approval controls	Limit or revoke a contract’s authority to spend tokens.	They do not reverse a direct transfer to the wrong recipient.

These defenses cover different failure modes. Address poisoning targets recipient selection. Malicious approvals target contract permissions. Red-pill attacks target a difference between preview and execution. Using one protection does not make the others unnecessary.

What to do when MetaMask shows an address warning

Do not confirm the transfer and do not let a counterparty rush you.
Take note of the network, address and warning category without exposing recovery credentials.
Close any untrusted site and reopen the official destination through a saved bookmark or independently verified link.
Obtain the recipient address again from the authoritative source.
Compare the full values character by character.
Search the address in the correct network’s reputable block explorer and review its activity. Explorer data provides context, not proof of ownership.
Contact the receiving platform through its official support page if the address is supposed to belong to it.
Proceed only after resolving why the warning appeared. When doubt remains, do not send.

If a legitimate address appears incorrectly classified, MetaMask provides a reporting path in some transaction alerts. Expand the alert, select the issue-reporting option where available and review the information being submitted. For other classifications, use official MetaMask support. Never share a Secret Recovery Phrase to appeal a warning.

What to do after sending to the wrong address

First, confirm the transaction hash, chain, asset, amount and destination using the correct block explorer. A pending transaction may sometimes be replaceable under network-specific conditions, but once a transfer is confirmed, MetaMask cannot reverse it.

If the destination belongs to an exchange or known service, contact its official support with the transaction hash. Recovery is not guaranteed and may involve a fee.
If the address belongs to an unknown attacker, avoid direct engagement and preserve evidence: transaction hashes, screenshots, domains, messages and timestamps.
Report the address through relevant wallet, explorer and law-enforcement channels for your jurisdiction.
If you also signed approvals or suspicious messages, review and revoke affected permissions. Moving remaining assets may be appropriate when account authority is at risk.
Do not pay “recovery agents” who promise guaranteed return of irreversible transfers or request your recovery phrase.

Sending to a wrong address does not automatically expose the sender’s private key. The next steps depend on whether the incident was only a mistaken transfer or included malicious signatures, compromised software or disclosure of credentials.

Clipboard-malware checks

If the address changes after copying, stop using the device for wallet activity. Disconnect it from sensitive accounts, update security software and investigate installed browser extensions, downloads and startup items. Use a known-clean device to protect important accounts where necessary.

Do not “test” suspected malware with a real wallet or valuable address. Reinstalling MetaMask alone may not remove system-level malware. If a Secret Recovery Phrase or private key was entered into an untrusted program, treat the wallet as compromised and migrate assets using a clean environment and newly generated wallet.

Practices that reduce address-poisoning risk

Never copy recipients from unsolicited incoming transactions.
Use an allowlist or address book only when entries were independently verified and changes require review.
Name known accounts clearly, but remember that labels are local conveniences rather than blockchain identity.
Separate daily dapp activity from long-term holdings and large transfers.
Keep MetaMask, the browser, operating system and hardware-wallet firmware current.
Remove unnecessary browser extensions and install software only from official sources.
Keep security alerts enabled unless you understand the protection and privacy consequences of disabling Basic Functionality.
Verify large payments with two people or a multi-signature policy when managing organizational funds.
Use transaction limits and withdrawal allowlists on services that provide them.
Pause when a recipient changes unexpectedly, even if the request comes from a familiar contact.

A two-person process for treasury transfers

Organizations should avoid making a single operator responsible for acquiring, entering and approving a destination. A simple separation-of-duties process can require one person to source the address, another to verify it through an independent channel and the authorized signer to compare the final wallet prompt.

Record the asset, chain, verified address, source of verification, test-transaction hash and final transaction hash. For recurring recipients, control address-book changes and periodically revalidate deposit instructions. Multisignature approval adds another review opportunity, although multiple signers can still approve the same poisoned address if they rely on the same shortened display.

Common misconceptions

“A dust transaction means my MetaMask was hacked”

Not necessarily. Anyone can generally send assets to a public address. The unsolicited activity becomes dangerous when it manipulates the user’s next action or links to a malicious site.

“The first and last six characters match, so it is safe”

No. Those are exactly the portions an attacker tries to match. Compare the complete address with an authoritative source.

“MetaMask showed no warning, so the recipient is legitimate”

No. New addresses and previously unseen campaigns may not yet carry threat intelligence. No alert is not a warranty.

“A hardware wallet prevents transfers to the wrong address”

A hardware wallet protects keys and provides a trusted signing display. It still signs the destination the user approves. Read the full destination on the device where possible instead of confirming automatically.

“The same EVM address works safely on every network”

The address format may be identical across EVM networks, but the receiving service may not support the selected chain or token. Confirm network support and deposit instructions separately.

“A successful test means I can paste again without checking”

No. A second paste can introduce a different address. Reuse the verified value through a controlled process and compare the final confirmation again.

Pre-transfer security checklist

I obtained the address from the receiver or service, not from transaction history.
I confirmed the exact asset and blockchain network.
I compared the complete recipient address.
I checked the final address displayed by MetaMask or the signing device.
I understand every warning and did not dismiss a malicious classification.
I confirmed whether a memo, tag or reference is required.
I checked the amount, token decimals and estimated network fee.
For a material amount, I completed a verified test transfer.
I am not acting under pressure from support, an investor, a romantic contact or a time-limited promise.
No one requested my Secret Recovery Phrase, private key or remote access.

Frequently asked questions

What is MetaMask address poisoning?

It is an attack in which a lookalike wallet address is placed in a user’s transaction history so the user may copy it during a later transfer. The fraudulent address often resembles a legitimate recipient at the beginning and end.

Can MetaMask block an address-poisoning transfer?

MetaMask can detect suspicious patterns and show recipient or security warnings, but users remain responsible for the final confirmation. Coverage depends on the network, wallet version and available threat intelligence.

What does “Malicious” mean in MetaMask?

It indicates strong signals that an address, token, website or transaction is associated with harmful activity. Do not proceed. Close the site and independently verify the destination through official channels.

Is a MetaMask “Warning” always a scam?

Not necessarily. A warning can reflect suspicious signals, missing verification or insufficient information. It means the request requires investigation—not that the alert should be ignored.

Can an attacker steal crypto by sending dust to my wallet?

The incoming dust alone does not authorize spending. Risk arises if the recipient copies the attacker’s address, follows a malicious link or signs a harmful transaction associated with the unsolicited asset.

Can I delete a poisoned transaction from MetaMask?

An onchain record cannot normally be deleted. Interfaces may hide spam assets or display warnings, but users should stop using history as the source for recipient addresses.

Does a verified badge guarantee an address is safe?

No. It signals confidence about recognized identity. It is not investment advice, a smart-contract audit or confirmation that the address is the destination you intended.

Does MetaMask share my recovery phrase with Blockaid?

MetaMask states that security checks do not share Secret Recovery Phrases, private keys or credentials with security providers. Relevant request data can be shared for simulation and threat analysis.

Should I send a test transaction every time?

It is most valuable for new destinations and material transfers. Consider network fees, minimum deposits and receiver requirements. A test complements full address verification; it does not replace it.

Can MetaMask recover crypto sent to a poisoned address?

No wallet interface can unilaterally reverse a confirmed blockchain transfer. A known custodial recipient may sometimes assist, but recovery from an unknown attacker’s address is unlikely and never guaranteed.

Final verdict

MetaMask address-poisoning detection adds a valuable warning at the moment a user selects a recipient. Its real value is creating a reason to stop before an irreversible transfer. It should be treated as one layer in a process that begins with an authoritative address source and ends with a complete comparison on the final confirmation screen.

The safest habit is simple but deliberate: do not use transaction history as an address book, do not trust shortened matches, verify the network and full recipient independently, and test consequential transfers. Security alerts improve that process; they do not replace it.

Official sources

Visit MetaMask ↗Official website