Welcome to the NAIP's English-language blog on IP-related news and topics for inventors, engineers, companies and patent professionals in Asia. As we get started, feel free to share your thoughts, comments or any suggestions, and check back often for the latest updates.

Bookmark this page! or Subscribe (RSS)

Novelty Grace Periods in Taiwan

by Jason Wang, NAIP Patent Engineering Research

In Taiwan, a patentable invention must satisfy three conditions: utility, novelty, and inventive step. Generally speaking, if an invention is known to the public before filing, it loses its novelty. However, under certain conditions, if the disclosure was made by the applicant, he can claim a "novelty grace period" in which the invention is still considered novel and patentable.

A novelty grace period is the allowable time from the disclosure date of the invention to the filing date. It's a special rule that, under a certain conditions, prevents the disclosure of an invention prior to the application to be recognized as prior art during examination. Many people assume the grace period is for the benefit of the inventor, allowing him to disclose his/her invention before filing an application. However, reality is opposite, because of the high risk of losing "novelty" in most or even all the patent offices if something unexpected happens during the period of the disclosure to application. Therefore, to help fully explain the details, here we will explain the grace period rules in TIPO and the risks that an applicant faces.

In the Taiwan Patent Act, Article 22, paragraph 2 deals with grace periods, and states:

an invention may not be patentable if it has been published, put to public use, or known to the public before applying for patent, unless it is filed within 6 months from the date of occurrence of the following causes:

  1. Where the invention is disclosed as a result of research or experiment;
  2. Where the invention has been exhibited at an exhibition sponsored or approved by the government; or
  3. Where the invention has been disclosed in an occasion not intended by the patent applicant.

Of the three causes, subparagraphs 2 and 3 are well-defined (though proving accidental disclosure is another matter), so here we focus on the first condition, paragraph 1.

TIPO's patent examination guidelines goes into further detail about this condition:

If the content of the invention sought to be patented has been published, put to public use, or known to the public prior because of the purpose of research or experiment, for example, the publication of a thesis, the novelty of the invention is kept if the patent is filed within 6 months from the date of occurrence, which is the date that the content of the invention is available in public, not necessarily read or known."

From the example provided in the examination guidelines, we have a rough idea that something like thesis disclosures satisfies the condition set under Article 22, paragraph 2, subparagraph 1.

Most recently on 11 December 2009, a draft amendment of the Patent Act passed to the Legislative Yuan for examination, and may take effect in the near future. In Article 22, "research" was removed from the "disclosed as a result of research or experiment" (subparagraph 1) and a second paragraph, "disclosed via printed publication", was added. According to the explanation from TIPO, "research" is something incomplete and requiring further discussion or study. An incomplete invention is not a subject matter of a patent. Therefore, "research" should not be included in consideration. The phrase "disclosure via printed publication", provides a wider range of the conditions acceptable for grace period status.

In the previous paragraphs we defined grace period and described the acceptable conditions under TIPO's Patent Act. However, one important fact must be mentioned: unlike in the United States, a grace period does NOT determine the priority date of the patent. Although TIPO accepts the patent applications that disclosed prior to filing and meet conditions set in Article 22, paragraph 2, there are 3 major problems/risks caused by this process:

(1) If the applicant claims the benefit of a grace period under Article 22, paragraph 2, subparagraph 1, the invention may lose patentability in some patent offices. For example, China's State Intellectual Property Office (SIPO) allows the grace period only if the disclosure is in a national or an officially-recognized international exhibition, or in a national-level scientific seminar. In the Japan Patent Office (JPO), although a similar grace period clause exists, the recognition of "disclosed as a result of experiment" is different from TIPO. So the risk of losing patentability in other countries/regions is a major concern.

(2) If the applicant claims the benefit of grace period under Article 22, paragraph 2, subparagraph 2, the "exhibition" in which the invention was disclosed plays a key factor in its patentability. Unfortunately, currently no set definition of "officially recognized international exhibitions" exists. But for example, at the EPO, only the World Expo has been listed as a qualified event for 2010. Therefore, while you can still file a Taiwan patent after an invention exhibited in CeBIT or COMPUTEX, it has already lost its "novelty" for a European patent at the moment of exhibition.

(3) There is still a risk of non-patentability even if the invention is disclosed under the qualified conditions of Article 22, paragraph 2, subparagraphs 1 and 2. If another inventor files a patent or discloses his own invention (with subject matter which can be used as prior art against your invention), during the time between your disclosure and your application, that creates a "lose-lose" situation. The other inventor cannot receive a patent because your disclosure has made his invention "non-novel". Also under the "first to file" system, his application or disclosure also invalidated your application because the grace period only allows a prior art exception of your own disclosure. No one is able to receive a patent in this situation—unless you or he can prove that the other achieved the invention by through illegal means.

Let's take an example to explain the problem an applicant may have, if he/she does not completely understand the differences between grace period and priority. Assume that Jenny is the inventor of a special controller for the PlayStation 3. She disclosed the controller and how it works at the E3 game show, on 16 June, 2008. She filed her invention on 1 February 2009--within the 1-year grace period offered by the USPTO. On 16 February 2009, Jenny decides to authorize her good friend Huang to produce this controller in Taiwan, so she requests a patent firm to file a Taiwan patent and claim the priority from the US patent. Do you see the problem here? Although the first filing date of the invention was only half a month before her decision to file a Taiwan patent (which satisfies the regulation for the international priority claiming) 7 months have passed since disclosure of the invention to the date of decision to file a Taiwan patent. The application has no chance to meet the requirements for the grace period in Taiwan. In this case, the controller may not be patentable in all patent offices except the USPTO.

To summarize, a novelty grace period is a remedial measure for filing a patent if you have already disclosed your invention. But is an option nonetheless. However, please don't assume that this rule erases any and all disclosure, because in many countries the invention may still lose novelty. The best way to secure the novelty of your invention is still to first file the patent. Then you can claim priority of the first filing date of when applying in other countries. Just ask yourself before disclosing an invention: "Is it worth the risk of losing this patent?" I believe you will know what to do next.


by Daniel Gross, NAIP Patent Engineer

In the almost anticlimactic wake of the Supreme Court's decision on the Bilski vs Kappos case, there seems to be little general consensus in the IP community as to what is patentable. One camp happily believes that by throwing out the machine-or-transformation (MOT) test as being definitive, the Court has further embraced the apparent 1952 Congressional intent of patentability of "anything under the sun that is made by man". Another camp views the results differently, thinking that the Court was actually contracting current scope of the statute and in fact merely added that it is not inconceivable for some future innovation to fail the MOT test and still be patentable. The Court, unfortunately, gave no examples of such an invention.

With this kind of legal ambiguity over what is patentable, many IP attorneys are gleefully lining their pockets, while, in an age of multi-multi-million dollar infringement lawsuits, the industries are scrambling to fortify against that suddenly-larger expense in product development and ownership known as uncertainty. Without such a clear example of what is patentable, even the USPTO appears hesitant to stray far from the MOT test, but I think they may be missing the Court's ultimate point. The way I interpret the decision, things might be a mess right now, but the writing is on the wall. The Supreme Court is carefully but surely slamming the door on the patentability of most software and business method patents as we know them by suggesting that they always have been, and still should be rejected as abstract ideas. Having no definitive MOT test just makes rejecting these "inventions" much easier.

In the Bilski decision, the Court threw out another bright-line test only to widen the concept of abstract ideas. With Bilski being the first, the Court also gave us a second example of the widened concept shortly afterwards when they vacated the Federal Circuit Court's decision in another patent case. The Lower Court had ruled in Prometheus Laboratories, Inc. v. Mayo Collaborative Services (Fed. Cir. 2010) that the claimed medical procedure was statutory because a drug induced modification of a patient's blood chemistry met the transformation fork of the MOT test. However, the Supreme Court directed the Lower Court to reconsider the case in light of the Bilski decision, suggesting that even meeting the MOT test is not enough to make a method statutory.

Most software and business method patents seen today would have been considered unstatutory material before the Federal Circuit's 1998 State Street decision, the case usually considered as opening the door for business method patents. Although ruled narrowly on the facts, the Bilski decision specifically singled out this same State Street decision as not being endorsed. "Nothing in today's opinion should be read as endorsing the Federal Circuit's past interpretations of §101. See, e.g., State Street, 49 F. 3d, at 1373."

Thus while many believe that the Justices were giving their de facto blessing to business method and software patents, the Court was actually doing exactly the reverse, implying that they were abstract ideas. Before Bilski, if a method passed the MOT test, it was automatically statutory material. After Bilski, this no longer proves to be the case. Why is a paper copy of software code not patentable, yet running the exact same software code on a computer suddenly turn an old machine into a new patentable machine? The Justices are asking the same kind of questions. Similarly, business methods can also be viewed as simply algorithms that use old machines, such as computers, phones, or websites to pass the MOT test, but still be abstract ideas at their core.

At the end of the decision, the opinion also included this line:
"In disapproving an exclusive machine-or-transformation test, this Court by no means desires to preclude the Federal Circuit's development of other limiting criteria that further the Patent Act's purposes and are not inconsistent with its text.”
This is spoken by a Court that wishes to contract rather than expand the scope of §101.

In response to the prodding by the Supreme Court in this decision, the Lower Court may take a different approach to developing the requested "limiting criteria" for patent eligibility. Perhaps instead of tackling the difficult job of defining what is patentable without running afoul of the law, it would be easier for the Lower Court to use what has already been long settled as statutory limits and merely define certain classes of subject material as "abstract ideas". To be sure, there would be countless court battles over whether a specific process falls into one of those classes, but there is no statutory problem with rejection of abstract ideas.

Defining classes of ideas as "abstract" could be relatively easy. For software: that which does not cause a "further technical effect" and which does not go beyond the inherent technical interactions between hardware and software. Similarly, for any process or method: that which does not cause a "further technical effect" and which does not go beyond the inherent technical interactions between the process and a machine. This basically is how Europe does it and universal symmetry of patent eligibility may go a long way toward easing some of the uncertainties of product development and ownership so prevalent today.

Unfortunately, US patent law lacks a specific "technical effect" requirement like that which is present in European patent law, so a "technical effect" requirement in the US might suffer the same fate as the MOT test (ruled as being inconsistent with the text of patent law). Nevertheless, the act of defining certain classes of ideas as abstract is not inconsistent with the Patent Act. It would merely be a generalization of historical interpretations of patentable subject matter and Supreme Court precedents such as Benson, Flook, and Diehr. In contrast, the recent push to expand patentability to software and business methods, however well intended, is in violation of the same historical precedents.

In the end, the key point is that the Supreme Court's rejection of Bilski's claims had nothing to do with the MOT test, and instead was specifically based on the rule that claims "can be rejected under the Court’s precedents on the unpatentability of abstract ideas". The Court even gave us a second example of the same rationale in the Prometheus case. The Lower Court was then basically challenged to find "limiting criteria" for patent eligibility. Again, the key point is "abstract ideas should be rejected".

While it is quite unlikely that the Congressional intent of patentability of "anything under the sun that is made by man" will be changed by this ruling, therein may be the solution. People "have" ideas, but "make" products--even digital ones. It may soon be up to applicants to prove that their software or business methods "made" something instead of being abstract ideas they "had" and then just written down or codified.


Back to Top