| Takeaway | Detail |
|---|---|
| USPTO reduced the 1(b) intent-to-use extension fee to $125 in 2025 | The 2025 fee reduction applies per extension request under Section 1(b), lowering the cost of delaying a statement of use for each mark |
| Multi-class filings multiply extension costs | Each class in a 1(b) application is charged its own extension fee, so a multi-class AI brand filing pays $125 per class per extension rather than a single flat fee |
| Verify the live fee before committing | USPTO fee schedules change; confirm the current $125 extension fee and per-class treatment on the USPTO's official fee schedule before filing |
| Compare like-for-like totals across filing strategies | When weighing multi-class applications against separate filings, total the $125 per-class extension fees plus base filing fees for identical class counts and terms |
The USPTO's 2025 reduction of the intent-to-use extension fee to $125 directly changes how multi-class AI brand portfolios should approach Section 1(b) filings, because the fee applies per class per extension.
Your total cost and strategy depend on the live USPTO fee schedule and how many classes you include, so verify the current option and compare like-for-like totals before committing.
How It Works
A Section 1(b) filing basis is used when an applicant files before the mark is in commercial use and registration follows an accepted Statement of Use. If use is not ready, the applicant can seek an extension of time to file the Statement of Use, but the current extension mechanism—including the live fee, whether it is charged per class or per request, and which classes are covered—must be confirmed on the USPTO's official fee schedule and the application record before any cost comparison. Do not treat a headline fee as an application-level or per-class charge until the live schedule confirms the charge structure.
As a worked check only, if the live USPTO fee schedule confirms a $125 per-class extension fee, one request covering a three-class portfolio would compute as $125 × 3 = $375, while a single-class filing over the same period would compute as $125; a second request would double those totals. Re-run the same multiplication with the live fee, the verified class count, and the number of extension periods actually needed before committing.
Do not rely on a remembered extension cap. Before scheduling, confirm the current maximum number of extension requests, the period each request covers, and the total time allowed from the notice of allowance in the USPTO's official rules and the application's TSDR record. Also verify the exact class list on each request before paying, because only the classes included on that request are covered by it.
Define the terms before you price anything:
| Term | What it means mechanically |
|---|---|
| Section 1(b) basis | Filing basis claiming bona fide intent to use; registration waits on an accepted Statement of Use. |
| Statement of Use (SOU) | The filing that supplies the verified date of first use in commerce for a class; registration issues after acceptance. |
| Extension request | The filing that buys additional time to submit the SOU; charged per class, per request, at $125 under the 2025 fee. |
Verify two things before committing. First, confirm the extension fee in the USPTO fee schedule under 37 CFR §2.6 on the date you file the request, since the schedule — not a summary of it — controls what is charged. Second, pull the application's class count and current notice-of-allowance deadline in the USPTO's Trademark Status and Document Retrieval system. Then compare like-for-like: equal class counts, equal numbers of extension periods, and the extension fee included in the total rather than quoted separately. A total that omits the per-class extension charge for the periods a class actually needs is not the complete option.
Key Factors to Consider
Three criteria decide whether a multi-class intent-to-use filing is worth committing to now: how many classes actually have use-ready goods or services, how many extension periods each class will need, and what the complete option costs when you add every per-class charge on the same fee-schedule date. Score each class separately before you look at a portfolio total — an AI brand portfolio often spreads across software, SaaS, and advisory services, and averaging those classes together hides the ones that will drive the extension bill.
Start with use-readiness. For every class, identify the specific goods or services, the specimen you can produce today, and the date of first use in commerce. A class with all three may not need an extension, while a class missing any one of them may be an extension-bearing class. Before multiplying anything, confirm on the current USPTO fee schedule whether the extension charge is assessed per class and how the official application record lists the classes.
Second, track the runway. Each extension request covers a defined period and is charged again for every class each time you file one, so the fee repeats rather than resolves. Pull the extension requests already filed, class by class, from your USPTO filing records, then compare that against the maximum period the rules allow. If the remaining runway is shorter than the time your product team needs to ship and document use, an extension is not a strategy; it is a delay with a price attached.
Third, build a like-for-like total. Compare one multi-class filing with extensions against splitting those classes into separate filings, holding three things constant: the fee-schedule date, the class list, and the number of extension periods. Then add the costs that scale with class count — attorney review per class, specimen collection, and the review time burned by any class that fails. A total that omits per-class professional time is not a comparison, it is a headline.
Treat every figure as a check, not a memory. On the day you file, open the USPTO fee schedule, capture the page, and rebuild the total line by line: base filing fee, per-class fees, and one extension charge for each class and each request. If a number on your worksheet cannot be traced to that page on that date, the option is not verified — and you are not ready to commit.
Common Mistakes
Pitfall one: treating a multi-class intent-to-use application as a single decision instead of a per-class one. Before relying on any fee reduction, confirm on the live USPTO fee schedule whether the extension charge is assessed per class; if it is, analyze each class individually. For an AI product that may span software, SaaS, and advertising-service classes, list every class on a one-page grid and mark whether you have a dated roadmap item showing use in commerce. If a class has no concrete launch milestone behind it, that class should not be included merely because the current per-class price looks attractive.
Pitfall two: comparing the extension cost against nothing at all — or against the wrong baseline. Readers routinely celebrate the lower per-extension price in isolation and skip the like-for-like total. The verify-before-you-commit rule applies here directly: before filing, pull the current fee schedule from the USPTO's fee page and compute, for each class you might claim, the full multi-year cost of a statement of use plus every extension you realistically expect to need. Do the same math for the fallback option — filing only the use-ready classes now and adding the rest in a separate application later. Then compare the two totals like-for-like, same classes, same timeline assumptions. If the deferred path totals less, defer.
A related mistake is treating the fee change as permanent policy. Fee schedules at the USPTO are set through rulemaking and can be revised, so the number you compute today is a snapshot, not a guarantee. The check: confirm the live fee on the official USPTO fee schedule before you commit, and note the date you checked it on your filing memo.
Finally, watch the reverse error: overcorrecting by dropping classes that are genuinely part of the brand plan. The fee cut changes the cost calculus, not the legal requirements — a bona fide intent must still exist for every class claimed. An AI startup that runs a chatbot, a developer API, and a professional-services line should keep each class that maps to a real planned offering, and cut only the aspirational ones. The rule: every class on the application must survive the question "can we show a credible plan to use this in commerce?" — and if the answer depends on the fee being low, the answer is no.
Insider Tactics
The inside move is not filing earlier or buying more time; it is separating the calendar you file against from the calendar you launch against. Teams that run a multi-class intent-to-use portfolio as a single launch plan tend to purchase extension runway in the classes that least need it and run short exactly where a specimen is hardest to produce. Two habits close that gap: split the portfolio along ownership lines, and anchor every extension request to an artifact you can already name. Both are verification exercises, not forecasting exercises, which is why they hold up when a product roadmap slips.
Start with ownership. When a class in an AI brand portfolio may be licensed, co-branded, or assigned to a subsidiary, an acquirer, or a regional partner, treat it as a candidate for its own application rather than one class inside the family filing. A multi-class application is a single proceeding, so carving one class out later means a division with its own fee and its own paperwork, and the carved-out registration carries the parent filing's history with it. Separate single-class applications filed on the same day preserve the same filing date while letting you move, license, or abandon one piece without amending the others. Before you assume splitting is the cheaper route, open the USPTO's current trademark fee schedule and price the division you would eventually need anyway.
The second tactic is to fund an extension only against a specimen you can point to. Before an extension request goes out for any class, write down the exact artifact you expect to submit — a live product page, a screenshot schedule tied to a release, a demonstration environment — plus the person accountable for it. If no artifact and no owner exist, the request is buying optimism rather than time. A named artifact also converts a portfolio-wide bet into a specific, checkable commitment that a reviewer who is not a trademark attorney can audit at a glance.
On timing, work backward from the launch milestone instead of forward from the filing date. Place the extension request on the calendar ahead of the docketed deadline with a buffer for the filing itself, and pull that deadline from the USPTO's Trademark Status & Document Retrieval record for each serial number rather than from an internal tracker, which drifts as personnel change. Re-check the fee on the day you file: the published fee schedule is the only authority for what you will actually be charged, and a figure you budgeted for a 2026 filing months ago is not evidence of the figure today.
Frequently Asked Questions
If my AI brand application covers multiple classes, do I pay one extension fee or several?
Each class in a Section 1(b) application is charged its own extension fee, so a multi-class filing pays the per-class rate for each class on every extension request rather than a single flat fee.
Does the fee reduction apply to every extension I file, or just the first one?
The reduced fee applies per extension request under Section 1(b), lowering the cost of delaying a statement of use each time for each mark.
When is Section 1(b) the right filing basis for an AI brand?
A Section 1(b) basis is used when an applicant files before the mark is in commercial use, and registration then follows an accepted Statement of Use.
How should I compare a multi-class application against filing separate applications?
Total the per-class extension fees plus base filing fees for identical class counts and terms so you compare like-for-like totals across filing strategies.
Could the extension fee change again after I plan my budget?
USPTO fee schedules change, so you should confirm the current extension fee and per-class treatment on the USPTO's official fee schedule before filing.
What drives my total cost most when filing an AI brand on an intent-to-use basis?
Your total cost depends on the live USPTO fee schedule and how many classes you include, since the extension fee applies per class per extension.
Quick answers
| What did the USPTO do to the 1(b) intent-to-use extension fee? | The USPTO reduced the 1(b) intent-to-use extension fee. |
| How is the reduced extension fee applied under Section 1(b)? | The fee reduction applies per extension request under Section 1(b). |
| How are extension fees charged in a multi-class 1(b) application? | Each class in a 1(b) application is charged its own extension fee, so a multi-class AI brand filing pays the reduced fee per class per extension rather than a single flat fee. |
| What should an applicant verify before committing to a filing strategy? | Verify the live fee before committing because USPTO fee schedules change; confirm the current extension fee and per-class treatment on the USPTO's official fee schedule before filing. |
| What should be totaled when comparing multi-class applications against separate filings? | Total the per-class extension fees plus base filing fees for identical class counts and terms. |
Also worth reading: Trademark Basics Explained Filing Your Brand At The USPTO: Trademark Basics Explained Filing Your · 7 Critical Trademark Filing Mistakes Startups Made in Early 2025 Analysis of USPTO Rejection Data: 7 Critical Trademark Filing Mistakes · 2024 USPTO Trademark Filing Fees A Detailed Cost Breakdown for Logo and Business Name Registration: 2024 USPTO Trademark Filing Fees