The block is the strongest remedy in the statute, and it works faster than an ordinary dispute. Under 15 USC 1681c-2(a), a credit bureau must block the reporting of information a consumer identifies as resulting from an alleged identity theft "not later than 4 business days" after it receives four things: appropriate proof of the consumer's identity, a copy of an identity theft report, the consumer's identification of the information, and a statement that the information does not relate to any transaction by the consumer. Subsection (b) then requires the bureau to notify the furnisher promptly that the information may result from identity theft, that a report has been filed, that a block has been requested, and of the block's effective dates.
A block and a dispute are different mechanisms with different standards. A dispute asks the bureau to reinvestigate accuracy; a block removes the information on the strength of the report and the statement. Under (c)(1) a bureau may decline or rescind a block only on three stated grounds: that it was made in error, that it rested on a material misrepresentation of fact by the consumer, or that the consumer obtained goods, services or money from the transaction. If a block is declined or rescinded the consumer must be notified promptly.
Three alerts exist, with different triggers and different durations. 15 USC 1681c-1 provides for each of them:
An initial fraud alert requires nothing but a good-faith suspicion, and it lasts "not less than 1 year" (subsection (a)). The ninety-day figure still in wide circulation is the superseded text: the 2018 amendments substituted "1 year" for "90 days", in the same act that created the free nationwide freeze. Any guidance still saying ninety days predates both. The bureau must also refer the alert to the other nationwide bureaus, so one request covers all three, and must give you a free file disclosure.
An extended fraud alert requires an identity theft report, and it runs for 7 years (subsection (b)). It additionally excludes you from prescreened credit and insurance offer lists for 5 years and entitles you to two free file copies in the following twelve months.
An active duty alert is for a deployed service member, lasts "not less than 12 months", and excludes you from prescreened lists for 2 years (subsection (c)).
An alert imposes a duty on the lender, which is what makes it more than a note. Under 1681c-1(h)(1), an initial or active duty alert notifies every prospective user of the report that the consumer does not authorize a new credit plan, an additional card on an existing account, or a credit limit increase, and no prospective user may do any of those things "unless the user utilizes reasonable policies and procedures to form a reasonable belief that the user knows the identity of the person making the request." There is a sharper version of that duty available on request: if the consumer specifies a telephone number for identity verification when placing the alert, a user must contact the consumer on that number before authorizing new credit (1681c-1(h)(1)(B)(ii)).
The statutory front door is IdentityTheft.gov. 15 USC 1681c-1(i)(6)(B) requires the Federal Trade Commission to establish a single webpage, within identitytheft.gov or a successor site, linking to each bureau's own page for requesting a freeze, the three alerts, and the prescreen opt-out. So the site is the mechanism Congress designated rather than a convenience someone built.