Tuesday, May 21, 2019

Should you ask your Attorney to perform a prior art search prior to filing a regular patent application?

Should you ask your Attorney to perform a prior art search prior to filing a regular patent application?

Upon asking several attorneys, it appears the answer is...

It depends.

Summary: An extensive prior art search undertaken prior to issuance of any patent in a patent family can increase the presumed validity of any patent of that family that cites said relevant prior art. All patents of a given family should cite all relevant prior art for the entire family prior to issuance. Early searching combined with crafting claims around such search results may significantly increase upfront search and attorney expense but a more precisely crafted claim set can maintain positive momentum through the patent office with an appropriately broad claim set.

However, an extensive prior art search in an operating company can expose the company to claims of willful infringement and triple damages. That can be mitigated by a freedom to operate letter, which are unreasonably complex and expensive for startups to consider. However, it can also be mitigated by a strong and valid offensive portfolio. Such a portfolio would be available for countersuit.

Professional Prior Art Searches:
* Using a third party searcher cost $1k-$2k
* Typically produce 24 references which can be culled by the attorney to 10 relevant references.
* Take as input a one or two paragraph summary of the invention, the heart of the invention, novel and important aspects. Search for patents which contain keywords of proposed independent claims. Search for patents with the assignee field equal to that of known competitors.

Reasons to perform a prior art search:
* Patent drafting and prosecution is a long and expensive process. Innovation and commercialization can be even longer and more expensive by one or two orders of magnitude. In the event of finding damaging prior art that would eliminate the potential to receive patent protection, it is better to avoid the expense and time of the filing of claims that will be invalidated or denied.

My opinion: Usually the go no-go decision will not rest on patentability of a particular claim, or in fact of any claims at all. However, raising external capital or doing an M&A deal is more challenging without a patent in place. Yet, a picture claim that issues in a patent may be sufficient for deals that are not IP-centric.

* Patent claim drafting can be focused to overcome known existing prior art.

* Patent claim drafting can be directed to non-obvious improvements to known prior art.

Reasons not to perform prior art search:
* No obligation on the inventor or their attorney to seek to discover prior art or to search prior art.
* An issued patent that is part of a patent family may be damaged by the discovery of prior art after the date of it's issuance. That prior art would be raised immediately in court as a challenge to the issued patent.
* Searching art and acknowledging it makes infringement thereof willful. Those triple damages can be mitigated by a freedom to operate legal opinion. However, the freedom to operate opinion is 100-200 pages of legal analysis.
* No need to search or disclose prior art in a provisional patent application or in a utility application. However, with a utility application an IDS would be filed citing prior art references.

How many references to cite?
* Once a patent issues, it is presumed valid over the cited references... with a 90% validity (clear and convincing) rather than a 50%+ (preponderance of evidence)...
* Usually submitting "too much" is better
* Examiner will initial references examined. Those are helpful in front of USPTO.
* Relevant references should be cited, even 100 or 1000. Anything that raises the issue of patentability.
* Irrelevant references can be viewed as an attempt to flood the system and discourage the examiner from evaluating them. Thus, it could weaken the standard of evidence later in court if fact finder perceives references were not relevant.
* Hundreds of relevant references are better.