First-to-File Pressure: The Cost of a Slow Invention Disclosure Pipeline

Priority is decided on a date, not on merit

The America Invents Act, Pub. L. No. 112-29 (2011), converted the U.S. patent system to first-inventor-to-file for applications with an effective filing date on or after March 16, 2013. Under 35 U.S.C. § 102 as amended, priority runs to the inventor with the earliest effective filing date, not to the inventor who conceived first or worked the longest.

That is the doctrinal reality that has been in place for over a decade. The operational reality has not always caught up. Many enterprise disclosure pipelines still treat the time between invention and filing as administrative drag, measured in weeks because that is how long the queue happens to be, not because that timeline reflects any priority calculus.

That treatment is misaligned with the statute. Every week a disclosure sits in intake is a week the universe of potential prior art grows. The internal disclosure pipeline is the priority pipeline.

What § 102 actually bars

Two subsections do most of the work.

Section 102(a)(1) bars patentability if the claimed invention “was patented, described in a printed publication, or in public use, on sale, or otherwise available to the public before the effective filing date of the claimed invention.” Anything the public could access, anywhere in the world, before the effective filing date is potential prior art.

Section 102(a)(2) bars patentability based on a U.S. patent or published application by another inventor that was “effectively filed” before the claimant’s effective filing date, even if not yet published when the claim is filed. A competitor who files the day before, in secret, becomes prior art the moment their application later publishes.

Both bars key to the effective filing date. Internal conception, internal disclosure, and internal review timelines do not move that date. Only filing does.

The grace period is narrower than counsel often briefs

Section 102(b)(1) preserves a one-year grace period for disclosures by the inventor or someone who got the subject matter from the inventor. The grace period is real, but it is narrower than its reputation.

It shields the inventor’s own disclosures. It does not shield against intervening third-party disclosures unless the inventor “publicly disclosed” the subject matter first. Most enterprise disclosure pipelines explicitly do not pre-publish. They file first. The default state is exposed.

The Federal Circuit narrowed the shield further in Sanho Corp. v. Kaijet Technology International Limited, Inc., No. 23-1336 (Fed. Cir. July 31, 2024). The inventor in Sanho had sold the invention privately before a third party’s intervening application. He argued the private sale qualified as a “public disclosure” sufficient to invoke the § 102(b)(2)(B) safe harbor against the intervening filing. The Federal Circuit held that “publicly disclosed” requires the invention to be made “reasonably available to the public”, and a private sale does not satisfy that standard.

The operational implication is direct. An inventor’s confidential pre-filing activity can simultaneously start an on-sale clock under § 102(a)(1) and fail to qualify as the public disclosure that would shield against intervening third-party filings under § 102(b)(2)(B). Both bars run during exactly the intake window most pipelines treat as benign administrative time.

The on-sale bar is wider than the language suggests

Section 102(a)(1) lists “on sale” as one of the prior-art categories. The Supreme Court resolved any doubt about scope in Helsinn Healthcare S.A. v. Teva Pharmaceuticals USA, Inc., 586 U.S. 123 (2019). A confidential sale to a third party, one with non-disclosure obligations, still places the invention “on sale” within the meaning of post-AIA § 102. The AIA’s catch-all “or otherwise available to the public” did not narrow the bar.

The two-prong test from Pfaff v. Wells Electronics, Inc., 525 U.S. 55 (1998), still controls: the invention must be the subject of a commercial offer for sale, and it must be ready for patenting. Both prongs can be satisfied months before anyone files an IDF.

The practical surface area is wider than most counsel brief to the business. Manufacturing supply agreements involving the invention. Beta deployments under confidentiality. Internal commercial offers between affiliated entities. Joint-development arrangements. Mark Lemley’s Ready for Patenting, 96 B.U. L. Rev. 1171 (2016), worked through how easily the Pfaff “ready for patenting” prong is satisfied during routine internal product maturation. Late disclosures often surface § 102 events that already started a one-year clock, sometimes a clock that has already expired.

The foreign-priority cliff

The 12-month foreign-priority window under 35 U.S.C. § 119 and Article 4 of the Paris Convention compounds the cost of intake delay.

The window runs from the U.S. effective filing date. Every week absorbed by internal review compresses the time available to evaluate, fund, and file foreign counterparts. That is operational pressure, but it is not the worst of the exposure.

The U.S. grace period does not travel. The European Patent Office, the Chinese National Intellectual Property Administration, and most other major jurisdictions operate on absolute novelty. A pre-filing public disclosure that is harmless in the U.S. under § 102(b)(1)(A) destroys novelty abroad. The same U.S. trade-show talk, conference paper, or product demo that the U.S. statute forgives is fatal in Munich, Beijing, and Seoul.

A disclosure pipeline that is slow at home is also a foreign-rights pipeline that is bleeding rights at the same rate.

Quantifying the cost of delay

The cost of delay is hard to quantify exactly because the prior-art universe is a moving target. The structure of the cost is visible enough.

The World Intellectual Property Organization’s World Intellectual Property Indicators 2024 reported that 3.55 million patent applications were filed worldwide in 2023, the fourth consecutive year of growth and the first year above 3.5 million. That is roughly 68,000 applications entering the prior-art universe every week, before counting publications, conference papers, and preprint posts. The probability that some subset overlaps a given disclosure rises with cycle time. In crowded fields (large language models, battery chemistries, next-generation power electronics), the probability rises faster.

Enterprise intake-to-filing pendency varies widely, but it is rarely measured in days. Practitioner reporting and the AIPLA Report of the Economic Survey consistently describe median time from invention disclosure to nonprovisional filing in the months, not the weeks. Each of those months is a month of accumulating intervening-art exposure that the grace period mostly does not shield against.

The operational frame for counsel is simple. Cycle time is a direct multiplier on intervening-prior-art probability. It is a direct subtractor from the foreign-filing decision window. It is a direct increase in the Pfaff-prong-two exposure surface for any internal commercial activity touching the invention. None of those costs show up on a filing-volume dashboard.

A real-world doctrinal trap

Sanho is the trap most worth memorizing for current practice.

The inventor in Sanho did everything that internal counsel might rationalize as low-risk: a pre-filing commercial relationship with a known business partner, under confidentiality. The Federal Circuit’s holding was that the relationship did not start the public-disclosure clock, and therefore did not invoke the safe harbor against an intervening third-party filing, even though it likely did start the on-sale clock.

The two clocks ran at the same time, in opposite directions. One disqualified the inventor’s claim through § 102(a)(1). The other did not save the inventor through § 102(b)(2)(B).

Both events happened months before anyone at the company likely thought of the work as “patent-relevant.” Both events were the kind of routine commercial activity that an active disclosure pipeline would have caught and triaged, and that a slow pipeline did not.

The pipeline IS the priority strategy

The consequence is operational, not philosophical. Reducing intake-to-filing cycle time is not an efficiency play. It is a substantive priority-preservation play.

Speed and thoroughness are not a tradeoff in a well-designed intake. A structured disclosure that captures the Pfaff-relevant facts (commercial offers, ready-for-patenting evidence, public-availability events) surfaces them at intake, when the company can still control the timing. The same structured disclosure surfaces the technical specifics that Berkheimer needs and the conception evidence the November 2025 USPTO inventorship guidance now centers. The intake layer is the leverage point for several distinct downstream decisions.

A disclosure pipeline that runs in days instead of months collapses the gap between invention and effective filing date. It collapses the foreign-priority decision window only modestly. It collapses the on-sale-bar exposure surface meaningfully. It collapses the third-party intervening-art window almost completely.

That is the case for the pipeline. None of it requires a doctrinal change. All of it requires intake that catches inventions earlier and packages them more completely.

What to read next

The doctrinal anchors here are stable. Helsinn on the post-AIA on-sale bar. Pfaff on ready-for-patenting. Sanho v. Kaijet on the narrow scope of “publicly disclosed” under the grace period. The MPEP § 2152 series for the examiner-facing version of § 102 prior-art categories.

For the structural problem these doctrines press on, see our note on the passive disclosure trap, the intake mode that produces the slow pipelines first-to-file punishes most.

ALID compresses intake-to-filing time by surfacing invention candidates directly from engineering artifacts, with source citations, before the standard IDF cycle runs. To see how the discovery and disclosure flow operates, read how ALID works, or request access to run a first discovery against your own engineering data.