The doctrine of secret prior art represents one of the more complex and conflicting areas of U.S. patent law.  In a process begun by the 1946 decision in Metallizing Engineering  and accelerated by Hobbes v. U.S.,  courts treat the effect on the validity of patents of sales, gifts, knowledge, public uses, private uses, advertising, and offers for sale all in very different ways, each dependent on a shifting and often co-extensive array of factors, most of which the USPTO is not equipped to evaluate when examining a patent application. As both the USPTO  and many patent attorneys  suggest, the new America Invents Act (AIA) language of 35 U.S.C. § 102 is intended to do away with secret prior art entirely.  It is also possible, however, that courts may instead take the language as an opportunity to harmonize the field.  

The secret prior art doctrine is rooted in important policy considerations: an inventor might keep an invention as a trade secret for a long period of time, only to quickly patent it when competitors start to unravel the secret, thus substantially extending the effective grant of a patent.   Seeking to avoid this result, pre-AIA 35 U.S.C. § 102 provided that “[a] person shall be entitled to a patent unless […] (b) the invention was […] in public use or on sale in this country, more than one year prior to the date of the application for patent[….]”.

Around the turn of the century and interpreting similar language, the Supreme Court considered the infamous corset case, Egbert v. Lippmann.   An inventor gifted a new set of corset ribs, or steels, to a woman who later became his wife.  Eight years later, he invited a friend to come examine the invention, whereupon his wife “went out, and returned with a pair of corsets and a pair of scissors, and ripped the corsets open and took out the steels. [The inventor] then explained to witness how they were made and used.”   Ten years after that, the inventor applied for a patent. 

The Court hinged its decision on the aspect of secrecy, regardless of whether sales, gifts, disclosure of knowledge, or uses were involved: “[i]f an inventor, having made his device, gives or sells it to another, to be used by the donee or vendee, without limitation or restriction, or injunction of secrecy, and it is so used, such use is public, even though the use and knowledge of the use may be confined to one person.”  

Sales parted ways from gifts, disclosures, and uses sixty years later in Metallizing Engineering  (Metallizing 2), an appeal from a district court case which laid out the underlying facts in greater detail (Metallizing 1).    An inventor discovered a new process for electrically roughing a hard metal surface, so that additional metal could bond more tightly to it, enabling a mechanic to rebuild worn engine parts.  Thirty months before filing for a patent, “he sought the advice of a friendly engineer who told him […] to test it out thoroughly in actual service before attempting to patent it.”   Twenty-nine months before filing, the inventor commissioned a patent search.   Twenty-three months before filing, the inventor placed an advertisement in a trade magazine, announcing that he had a new process for metalizing hardened metals, though no orders came of the notices.  The inventor also spoke with an executive who “offered him a contract whereby the inventor was to apply for a patent and give an exclusive license thereunder.”  Between thirty and twelve months of filing, the inventor solicited and received sixty-six orders for rebuilt parts, although “the practice of the process was so guarded as not to come to public knowledge; its nature was disclosed only to a few employees and advisers of the inventor, less than half a dozen in number, in all cases under a promise of confidence which was not abused.”   An undetermined number of these jobs were done at the inventor’s own expense,  essentially becoming gifts, and other sales were to parties who knew that a ‘secret process’ was being used.  One order, eight months before the patent filing, was performed for customer who understood that the job was experimental. 

Relying in part on Egbert, the court of Metallizing 1 decided that the foregoing sales and uses were not public, and thus not a bar to patentability.  Metallizing 2 reversed, led by Judge Learned Hand in writing that “it is a condition upon an inventor's right to a patent that he shall not exploit his discovery competitively after it is ready for patenting; he must content himself with either secrecy, or legal monopoly.”   Importantly for later doctrine coherence, however, the court never identified exactly which activities gave rise to the patentability bar.  Perhaps the defining event was the discussion with the engineer, or the advertisements.  Perhaps the bar was triggered by the gifted parts, or perhaps only those sales which resulted in making use of the inventor’s secret to gain “a competitive advantage over others”  before patenting, although Metallizing 1 made no findings as to exactly which sales these might be, and Metallizing 2 proffered no factors by which to judge competitive advantage.

Thirty years later, Hobbs v. U.S.  clarified that sales in general -- regardless of purpose, conditions of secrecy, or identity of the seller -- gave rise to a 35 U.S.C. § 102 ‘on sale’ patentability bar, so long as the device was sufficiently developed for patenting and experimental use had ended.  But determining whether and when a contract rendered an invention ‘on sale,’ whether inquiries or orders counted as sales, and whether and when experimental use had ended, all proved no quick task, even in Hobbes.

Complexities soon mounted.  Consider a bare handful of the relevant doctrines, all of which are citable law: sales themselves may be experimental uses and hence exempt from the § 102 ‘on sale’ patentability bar;   the sale of rights does not place a device ‘on sale’;   sales between legally related entities may not be sales;   an unreceived or rejected offer for sale may bar a patent;   no delivery or payment is necessary;   neither the inventor nor the end user need know of or consent to the sale;   an entity may take orders before the device is perfected if no samples are given to customers;   but ‘makeshift’ models shown to potential buyers may bar patentability;   except that describing the fine details of an invention to a potential buyer before a sale negotiation may not;    likewise the disposal of prototypes which results in income may not;   the preparation of price lists may suffice to put an invention on sale;   and experimental uses by third parties may trigger the fatal bar, unless under the supervision and control of the patent applicant. 

The doctrinal multiplication has side effects: an independent inventor looking for investors or customers before seeking a patent – itself an increasingly expensive exercise – is likely to run afoul of at least one ‘on sale’ exception.  Sophisticated inventors and large inventive entities are better able to navigate these treacherous waters, while courts must deal with intensely fact specific sales situations and oft-conflicting case law scattered across the better part of a century.  

Perhaps appreciating this, the America Invents Act amends 35 U.S.C. § 102, which now states “[a] person shall be entitled to a patent unless […] the claimed invention was […] in public use, on sale, or otherwise available to the public [….]”   According to many readings, the ‘or otherwise available to the public’ language modifies both the ‘use’ and the ‘on sale’ clauses, doing away with ability of secret prior art to form a bar to patenting.  It remains to be seen whether courts will adopt such an analysis, as patents filed under the America Invents Act have typically not yet issued, let alone been challenged in court. 

But there is an option beyond the wholesale binning of secret prior art, if ‘available to the public’ is read into both the ‘use’ and the ‘on sale’ clauses: sales could simply be aligned with the current handling of uses, gifts, and other disclosures.  Pre-AIA § 102 case law already requires that in order for a non-economic use (such as a gift or demonstration) to qualify as a public use, it must be accessible to the public.   Adequate secrecy conditions may prevent gifts, disclosures, and uses from becoming a pre-AIA § 102 ‘public use’ bar  – just as a lack of security rendered the wearing of corset steels a public use.   Indeed, the Federal Circuit may be moving in this direction for questionable ‘on sale’ situations already.    

In Delano Farms,  decided post-AIA, but interpreting a patent filed pre-AIA and applying pre-AIA law, a third party obtained control over test varieties of improved grapes under a verbal condition of secrecy.  The third party, however, then gifted the grape plants to two others over four years, all of whom were aware that the improved grapes were a major competitive advantage, and all of whom grew hundreds of the plants in locations that were visible from public roads.  

By most measures, had any of these transfers or activities been offers for sale, they would have rendered the plants unpatentable under Metallizing.  Evidence to this effect existed: as the lower court noted, one of the third parties marketed the others’ grapes and acted as a mentor,  and checks for ‘grafting’ were exchanged between the parties.  The parties may have exchanged social or non-monetary favors for the plants, and likely distributed the plants among one another with an understanding as to the future sale of the grapes.   The trial briefs and complaints variously claim that the third parties grafted the grapes in the ordinary course of business  and that the patents were invalid under 35 U.S.C. § 102(b)  – though without specifying whether the plaintiffs wished to apply the ‘public use’ or ‘on sale’ bar.  A mere conditional offer to sell the grapes if the harvest was good, perhaps combined with a few sample grapes, should have sufficed to invoke the later.  

But the question of whether the gifts were essentially a sale was not addressed.  Rather, the Federal Circuit decided the matter under the ‘public use’ doctrine, applying a ‘sufficiency of secrecy’ test very reminiscent of Egbert and its progeny.  The court cited factors such as the inability to identify grape varieties by viewing the vines alone, the limited number of plantings as compared to the size of the field, the failure to label the plantings, the lack of evidence that any other person had ever recognized the test varieties, and most especially the third-parties’ apparent treatment of possession of the test vines as confidential.   

Functionally then, the application of a post-AIA  § 102 ‘[publically] on sale’ bar might look quite a lot like Delano Farms.  A court’s decision would hinge on whether any sale or offer was ‘publically available’ – that is, made to a party not under a sufficient obligation of secrecy – as well as whether there was a publically-available use.  Under such a regime, a backyard inventor need not find his patent invalidated by a few tightly-restricted sales of samples: at least, provided those secret sales did not give rise to new art, reverse engineering, or publically-available uses more than one year before the patent filing date.  Like secret uses, secret sales could become publically available sales when distributed too broadly.  

Importantly for courts reviewing such cases, the record then becomes a matter of public availability and the measures parties took to maintain secrecy.  Courts could turn to the business of deciding what manner of “limitation or restriction, or injunction of secrecy”  suffices to prevent a sale from becoming ‘available to the public,’ rather than the tribulations of sorting a hodgepodge of old sales receipts and oral testimony.

Finally, Delano Farms should provide a measure of reassurance to courts who are troubled by the potential implications of post AIA 35 U.S.C. § 102 – reviewing gifts under the somewhat more rational ‘publically accessible’ standard, rather than the Metallizing rule of sales, has not resulted in any great parade of horribles.   Subjecting offers and sales to a similar scrutiny should, at the least, do no worse.