The AIA’s First-to-File Transition SHOULD have Resulted in More Provisional Filings

best-practices-businessmanThe United States has stood apart as one of the few countries that did not abide by the first-to-file system until recently. Specifically, the patent statutes previously identified novelty and non-obviousness requirements made with respect to the “time of invention”. The American Invents Act (AIA), which was signed into law five years ago, changed the focus to what was prior art as of the “effective filing date”. Thus, the AIA transitioned patenting in the United States to become a “race to the patent office”.

Many had argued that the first-to-invent was a trap for the unwary as the second-to-file was almost certain to lose in an interference proceeding convened to determine who was first-to-invent, yet applicants fell in love with the idea that they could take their time to file their patent application. Under the AIA, the second-to-file can be the inventor if the first-to-file is challenged successfully in a derivation proceeding, but there have been few of those administrative determinations and the odds of success seem to be poor based upon the thin data so far.

A prominent advantage of the first-to-file system is that it can simplify determinations as to what references can be applied as prior art and – when multiple parties are attempting to patent very similar ideas – which party is entitled to a patent (upon clearing the rest of the patent requirements). However, a prominent disadvantage is that there is rarely a clearly defined “invention date” on which an invention has been completely, or even sufficiently, identified. Frequently, after parts of an invention are initially identified, further testing and/or assessment are required to identify enabling details, preferred embodiments, and/or concepts of highest business priority.

The first-to-file timeline leaves applicants in a quandary: Should a patent application be filed as soon as an idea is formulated, or should the filing be postponed until sufficient consideration and development has occurred so as to identify the pertinent and valuable details of the invention? Filing an application too soon may prevent an applicant from modifying or supplementing a disclosure with new results or implementation specifics, whereas filing an application too late may prevent an applicant from securing a patent due to a potential intervening filing.

Fortunately, the United States preserved a vehicle that would allow an applicant to manage these risks: the provisional patent application. A filing date of a provisional filing date may serve as an “effective filing date” of a non-provisional patent application that claims priority to the provisional (so long as what is being claimed in the non-provisional application was disclosed in the provisional application). Meanwhile, an applicant may further supplement a provisional’s disclosure and/or select amongst embodiments in the non-provisional application. Thus, the provisional can provide the applicant with a tool for putting a “stake in the ground” with respect to an invention while allowing an applicant to further research the invention and/or consider business priorities with respect to the invention. These advantages supported previous recommendations to “File first. File often. in the post-AIA world, such that applicants could routinely establish their first-to-file status. Many of the authors’ clients file provisionals early and often to win that race to the patent office.

Have applicants generally been following this strategy? To examine this question, we requested data from the United States Patent and Trademark Office for the number of (1) provisional and (2) utility non-provisional U.S. patent applications where the first inventor was a resident of the United States. We focused on applications with first inventors residing in the United States both because we felt that the data approximated decisions as to whether precipitate a United States non-provisional filing with a provisional (e.g., so as to attempt to exclude circumstances that, in effect, prevent such priority filings due to a foreign priority filing) and because the data was rather readily available.

Figure 1 shows the number of provisional (green triangles) and utility filings (blue squares) per fiscal year. Both filing types reached the maximum counts in fiscal year 2013 (October 1, 2012 through September 30, 2013). This fiscal year included the transition to the first-to-file system, which occurred on March 16, 2013 (vertical dashed line): for applications with effective filing dates on or after March 16, 2013, prior art was to be identified based on effective filing dates of the applications rather than invention dates. Notably, the date threshold was to be assessed based on an application’s effective filing date. Thus, either provisional or non-provisional applications filed before this date could, not only themselves escape the AIA prior-art definitions, but can also serve as a priority application to provide subsequently filed applications with an “effective filing date” under the old system. Therefore, the uptick in provisional filings during fiscal year 2013 may not be surprising.

FIG 1-2

Figure 1.

Since 2013, the number of provisional filings has decreased. However, this trend in the number of filings need not reflect applicant prioritization of provisional applications. Absolute numbers of filings may depend on many variables, such as patents’ perceived value, the economy, the perceived probability of a patent application being allowed, and so on. Consistent with a potential impact of one or more other factors, we see that the number of utility filings also decreased since 2013.

Thus, to investigate whether applicants have been adopting a strategy of filing more provisional applications, we determined a ratio of the number of provisional filings (with a first-named inventor identified as being a resident of the U.S.) relative to the number of utility filings (with a first-named inventor identified as being a resident of the U.S.). As shown in Figure 2, this ratio also exhibited an uptick in fiscal year 2013, but the ratio has since returned to Fiscal-Year-2012 levels. Thus, this data suggests that applicants have generally not changed their filing strategies to file more provisional applications in view of the AIA’s change to define prior art based on applications’ filing dates.

FIG 2-2

Figure 2.

Why have applicants not changed their strategy to file more provisional applications? The answer to this question may depend on technology areas. For example, with respect to biotechnology, provisional applications have been traditionally frequently used, as they provided an opportunity to extend an expiration of a corresponding non-provisional application without affecting patent term. As another example, many software applications were, and unfortunately still often are, filed without thoroughly disclosing implementation details. Such high-level disclosure may not require time investment sufficiently long to warrant any delay in a non-provisional filing. Some practitioners have been skeptical of provisionals, citing their informality as their biggest flaw given that the filing may not capture the ideas properly in the rush to get a fling date.

5th-anniversary-200While explanations may be provided for individual technology areas, applicants and/or ideas, the authors nonetheless believe that our data indicates that applicants are not strategically using the opportunity provided by the provisional application. This type of application allows an applicant to secure a priority date for prior-art purposes and clearly indicate possession of ideas. Both of these advantages are all the more important in the wake of the AIA’s transition from a “time-of-invention” to an “effective-filing-date” focus. Further, provisional applications provide unique prosecution-strategy flexibility, as they are not published (until they are included in a priority chain of a published non-provisional application), need not be included in a priority chain, and are associated with inexpensive government fees. Admittedly, it is important that provisional applications be prudently drafted so as to fully support a subsequently filed non-provisional application, so drafting costs are not insignificant. However, an applicant may build on a disclosure of a provisional for a non-provisional filing, such that part or all of the drafting costs can, in effect, be recouped.

Therefore, the authors believe that applicants and practitioners should begin reassessing their drafting-and-filing workflows and consider utilizing provisional applications to improve allowance probabilities (by reducing the number of references that can be used as prior art), increasing the likelihood that the ideas they invent first are not attributed to others and improve the value of patents to the applicant (e.g., by providing applicants with extra time to identify valuable claim focuses). “File early—file often.”


Warning & Disclaimer: The pages, articles and comments on do not constitute legal advice, nor do they create any attorney-client relationship. The articles published express the personal opinion and views of the author as of the time of publication and should not be attributed to the author’s employer, clients or the sponsors of

Join the Discussion

6 comments so far.

  • [Avatar for Dave]
    September 19, 2016 03:54 pm

    Good article. I’d be interested in seeing an analysis of what percentage of nonprovisionals claim priority to at least one provisional application. It seems like that would be helpful data, as it would answer the related question of whether applicants are more often taking advantage of the provisional application process (i.e., not just whether there are more or fewer provisionals as compared to nonprovisionals).

  • [Avatar for Anon]
    September 15, 2016 05:23 pm


    I do not have the stats on hand, but at least in the US, the dividend from innovation from the small firms FAR outweigh the dividends of the large firms.

    That is, when the large firms even let the dividends be captured in the US system and are not too busy playing offshore shell games.

  • [Avatar for Edward Heller]
    Edward Heller
    September 15, 2016 04:47 pm

    Benny, big companies (other than drug companies) tend to invent iterations. Small inventors and startups tend to invent entirely new products and technology.

    There is a difference.

  • [Avatar for Benny]
    September 15, 2016 04:35 pm

    Edward, lose on the swings, gain on the roundabouts. Who has more impact on the quality of your life, the small inventors or the big internationals? Whose patents usually disclose more innovation.? Which of the two alternative policies serves the greater good?

  • [Avatar for Anon]
    September 14, 2016 04:05 pm

    Mr. Heller, I agree with your sentiments here.

    I would only add (as I have remarked in this forum previously), one of my near constant client education efforts concerns the level of work necessary for a good provisional application is every bit on par with a good non-provisional application. Far too many out there mislead inventors as to that level of effort, and I am constantly re-educating new clients that no, you should not (for example) simply file your marketing materials as a provisional application.

  • [Avatar for Edward Heller]
    Edward Heller
    September 14, 2016 03:04 pm

    Gotta love these false narratives about first-to-invent. Everyone really knew that interferences were normally won by the first to file. But interferences were between rival applicants to the very same invention. The primary benefit of first-to-invent was the ability, for one year, for “US inventors” to swear behind prior art by showing prior invention of that prior art. It was this benefit more than anything else that forestalled any need for a race to the patent office, and it allowed American inventors to continue to develop their inventions with some degree confidence that some sudden disclosure by someone else would end their investments in their invention and their dreams of success.

    A provisional filing strategy does offset this problem to some degree. But the claims in the non provisional must find 112(a) support in the provisional. This is different from removing prior art that one invented first.

    To understand this distinction, consider the discovery of species A of genus Z by inventor 1. A provisional application disclosing A will support a non provisional application claiming species A, but it would not support a non-provisional application claiming genus Z. Thus the intervening disclosure by party 2 of species A would prevent the patenting of genus Z. However, in first to invent, the prior invention of species A by inventor 1 could be used to “remove” a reference disclosing species A so that the inventor could claim not only species A in the non-provisional, but genus Z.

    Obviously, America shot itself in the foot if not in the head when I went to first-to-file. It is not a good system as compared to the first-to-invent system that we abandoned.

    The five-year anniversary of the AIA is a time to mourn the passing of a great patent system. The only beneficiaries were non US companies, or companies whose businesses were truly international so that they needed clearance worldwide in order to market their products. The people harmed were American inventors, American startups and American universities.