The Defensive Patent License

JENNIFER M. URBAN JASON M. SCHULTZ

BERKELEY LAW

2013 O’Reilly Open Source Conference Portland, Oregon

July 24, 2013

PROBLEM

Innovation in the shadow of software patents* Especially for Open Innovation Communities FOSS

Open Hardware

Mixed- or Sole-Strategy Companies

Also a problem for non-OIC tech companies

*Overbreadth and other quality problems with software patents

POSSIBLE SOLUTION: DEFENSIVE PATENTING

“If you can’t beat ‘em…” (Or, “If we’re getting patents anyway…”)

Seek patents to deter offensive lawsuits, not for licensing or exclusion purposes In theory, only asserted in response to litigation

In patent-heavy industries with well-resourced competitors, often leads to cross-licensing détente Requires some level of overlapping technology

Dealing with software patents, specifically: “your broad, vague patents can be matched by mine”

THREE POTENTIAL USES FOR DEFENSIVE PATENTING Patent threats between competitors Patent threats from bullies Patent threats from trolls

COMPETITOR THREATS Patent threats between competitors can prove costly High information costs to avoid inadvertent infringement

•  Too many patents, too many claims, valuation problems •  Fear of willfulness discourages patent mapping

High damages

High costs of injunctive relief

•  Less after eBay v. MercExchange High cost of litigating

HOW DP HELPS Defensive patenting ! cross-licensed portfolios

•  Addresses information costs •  Limits threats of injunctive relief •  Limits litigation costs

Creates greater certainty and predictability for innovation cycles, market segmentation, and risk profiling which leads to greater investment, R&D, and increased consumer welfare

WHAT ABOUT BULLIES? Patent threats by Patent Bullies (Goliath v. David)

•  Verzion, AT&T and Sprint v. Vonage (VOIP) •  Apple v. HTC (smartphone touch screen) •  Microsoft v. Linux (operating system features) •  Blackboard v. Desire2Learn (educational CMS)

Defensive patenting could help reduce threats and costs but for several reasons has been underutilized

WHY DON’T OSS PROJECTS PATENT DEFENSIVELY?

You all know best: more reasons, please! But these are a few from our work with a variety of open innovators: Current model requires concentrated costs and benefits

•  Legal, information/search, transaction costs to build portfolio and fight back against threats

•  Start-ups and F/OSS projects generally lack these resources; but see OIN (Oracle, IBM, NEC, Novell, Philips, Red Hat, and Sony)

(Understandable) Cultural and political opposition to software patents

•  Used by bad actors: anti-competitive, bullying, trolling •  Inappropriate SM: patenting math, laws of nature •  Recent lawsuits not building confidence in the patent system…

(Understandable) Mistrust of “defensive” commitments and longevity

THE DPL Distributed network of defensive patents Distributed license structure similar to FOSS/CC Licensor offers [entire] patent portfolio under DPL

•  Automatic NE, RF, perpetual, worldwide licenses to all comers who also commit to offering their portfolios under the DPL

•  Irrevocable unless (1) licensee sues DPL user non-defensively or (2) stops offering own patents under DPL

DPL users may continue to license or litigate against parties outside the network (those not using the DPL) DPL users may stop offering DPL with appropriate notice

•  We are currently considering six months as appropriate •  Previously issued licenses to other DPL users remain in effect after

a party leaves •  But see Hayes/Schulman “sticky”/”non-sticky” idea

•  Leaving users’ DPLs may be revoked at each licensor’s discretion DPL obligations “travel with the patent”

HOW IT WORKS: EXAMPLE OF THE DPL OVER TIME

DPL OVER TIME

DPL OVER TIME

DPL OVER TIME

DPL OVER TIME

DPL OVER TIME

HOW THE DPL COULD HELP Distributes both costs and benefits via network effects

•  Costs of patenting are distributed to those who can afford them or already need to patent; possible pro bono help

•  All users benefit from each commitment to defense

•  Eliminates info, injunction, litigation costs w/in network

•  Costs of cross-licensing are low, as with FOSS and CC licenses

Uses high legal, information, and transaction costs of patent strategy to encourage growth of network

•  High info costs & risks vs. commitment to defense and associated certainty may incent new firms to join network, increasing its value

Provides legally binding commitment to defense: multi-lateral disarmament

•  Licensees can stop worrying about particular patents coming back to haunt them, even if ownership changes

Promotes “Safer Patenting”

Supports Open Innovation Community cultural and political norms

•  IP used to promote freedom to innovate, access to knowledge, and protection from legal constraint

•  Like instrumentalist approaches of FOSS and CC in using copyright

Technology Neutral

Can be done , entirely through private choice

•  No changes in examination, PSM, claiming needed

•  But can work along with these solutions

Co-exists peacefully with other solutions

WHAT ABOUT TROLLS?

Defensive patenting generally ineffective because trolls do not make, use, sell, or offer for sale, so you can’t assert a patent against them But a true, permanently defensive patent has little value to trolls: safer patenting!

http://www.flickr.com/photos/thomwatson/

The Fremont Troll

WHAT ABOUT PATENT QUALITY? Patents are the best prior art: DPL can help build it out DPL users should also support other ideas for increasing patent quality and clarity (Menell, Lemley, Chin, others) Encouraging networked defensive patenting is indeed a second-best solution Can absolutely work in connection with other ideas for reconsidering PSM, tweaking examination, etc

WHAT ABOUT PATENTABLE SUBJECT MATTER AND CLAIMING?

CONCERNS Need overall patent system reform

Patentable subject matter Patent overbreadth and vagueness

Technical and Practical Concerns: Most Addressed; Some a Matter of Strategic Choice

•  Network has to hit a size threshold for some benefits to kick in •  Weak incentives to patent and/or join •  Adds to patent thicket •  No requirement of mutual defense •  Too much to ask for whole portfolio [could be by standard or

technology •  Too hard to leave •  Too easy to leave •  Antitrust concerns, especially in EU •  Gaming the system/free riding/bankruptcy

NEXT STEPS Presently really “kicking the tires”

Terms

Language

Gathering adopters

•  Presently discussing terms and language with interested OS projects, start-ups, medium-portfolio companies

Launch planned for end of 2013

THANKS! Jennifer: jurban@law.berkeley.edu

Jason** SchultzJ@exchange.law.nyu.edu **contact for comments and sign-on discussions

-------------------------------------------------------------------------------- GitHub: https://github.com/defensivepatent/model-defensive-patent Paper: http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2040945 Op-ed: http://www.wired.com/opinion/2013/02/nuclear-deterrence-for-patents-lets-create-a-network-of-defensive-patents/ David Hayes and Eric Schulman response: http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2054314