There are NO intellectual property rights issues with Open XML

One of the interesting pieces of FUD floating around out there about Open XML is that there are IP issues with it and that is why countries should reconsider yes votes etc. etc. This is a factually wrong line of logic, and one that has been perpetuated throughout the process because most people a) don't have a law degree and b) don't want to take the time to think about - so it is easier to assume the worst.

The following text comes from a piece that was pulled together to help people understand these issues. It is too long of a doc to put up in its entirety, so I'm just going to pull a few sections.


Microsoft has made legal commitments to Ecma International, to ISO/IEC, and to all interested users and vendors that anyone can use and implement Open XML without IPR burdens. Microsoft believes that it is in everyone’s interest for this open file format to be available freely and easily for document exchange and preservation. When Microsoft submitted and turned over control of Open XML to the international standardization process, Microsoft also provided multiple options to ensure that its essential patents can be used by anyone, including OSS developers. These IPR commitments go beyond the requirements for ISO/IEC adoption of a standard, and ISO/IEC and Ecma have stated specifically that there are no IPR issues with Open XML.

Any Required Microsoft Patent Rights Are Available On A Royalty-Free, Perpetual Basis To All Implementers, And Both ISO/IEC And Ecma Have Publicly Declared that No IPR Issues Exist.

Microsoft made a patent declaration to Ecma and agreed to make any of its patents covering Open XML available consistent with Ecma’s “Code of Conduct for Patent Matters.” (See (Sec. 2.2) and

Microsoft also submitted to ISO/IEC a “Patent Statement and Licensing Declaration Form.” The ISO/IEC form provides three checkboxes: (a) willing to license necessary patent claims on RAND-Z (royalty-free) terms, (b) willing to license necessary claims on RAND (royalty-bearing) terms, and (c) unwilling to license necessary claims under (a) or (b). (See this link) Microsoft checked the first box. That means that if someone asks for a RAND-Z license to implement Open XML, we must provide such a license.

Microsoft also attached to its ISO/IEC patent declaration a commitment that implementers of Open XML would have the benefit of our “Open Specification Promise” (OSP -- available in Appendix A and at and our “Covenant Not to Sue” (CNS -- available in Appendix B and at as an alternative, if they prefer.

Microsoft thus has gone much further than what Ecma and ISO/IEC require. Both require that a company offer to license its necessary patent claims on RAND terms (which could include a royalty). Microsoft has instead offered all implementers their choice between (a) a negotiated RAND-Z license, (b) the OSP, or (c) the CNS, all three of which provide for royalty-free use of Microsoft’s necessary patent claims.

Indeed, Ecma and ISO/IEC have publicly stated that there are no IPR concerns with Open XML. In a document explaining the upcoming Ballot Resolution Meeting (BRM), ISO/IEC noted that IPR issues will not be discussed, because, “IPR decisions have previously been delegated by all the ISO/IEC and IEC members (NBs) to the CEOs of IEC and ISO/IEC, and they in turn have examined them and found no outstanding problems.” (emphasis added). Ecma issued a similar statement. (Sec. 2.2).

To recap, because the BRM and comment process is designed to ensure that the specification is fully and correctly defined, and because ISO/IEC has found no outstanding IP issues, there are no IPR issues associated with Open XML that should raise concerns about implementation, long-term document retention, preservation, or accessibility.

Adoption of Open XML -- Including By the Open Source Community -- is Growing Exponentially, Underscoring that Developers and Customers are Comfortable that there Are No IPR Issues with Open XML.

Thousands of developers, organizations, governments, and professionals spanning 67 countries and six continents have already expressed public support for Open XML and for its approval by ISO/IEC. (See and More than 2,000 members have joined, and hundreds of independent software vendors are developing solutions using Open XML.

A growing number of implementations of Open XML -- including from open source developers -- are becoming available, including those released by Apple (Mac OS X Leopard, iWork 08, iPhone), Adobe (InDesign), Novell (SUSE Open Office), Microsoft (Office 2007, Office 2003, Office XP, Office 2000), Mindjet (MindManager), Palm Intergen, OpenText (LiveLink), Dataviz (DocumentsToGo on Palm OS), NeoOffice, and Altova (XMLSpy), as well as those under development by Corel (WordPerfect), Gnome (GNumeric), Xandros, Linspire, Turbolinux, and others.  These implementations are now available on many platforms, including Linux, Macintosh, Windows, Java, .NET, and handheld devices (PalmOS, Symbian, Windows Mobile).

Key Aspects Of Microsoft’s OSP

  • Any required Microsoft patent rights are freely available to all developers and customers of
    Open XML in either open source software or proprietary software.
  • By stating that the covenant is “irrevocable,” Microsoft has assured users that there will not be a change in company policy at any point in the future.
  • Vendors, distributors, and users of Open XML implementations benefit from the OSP just like implementers do. Consequently, there is no need for implementers to pass the promise on to others in their distribution channel, as it is always available to everyone directly.
  • No one needs to sign anything or even reference Microsoft to take advantage of the OSP.
  • This form of patent non-assert enables open source software implementations. It is especially convenient for open source software developers as there is no issue as to whether or not the IP is sub-licenseable.
  • The OSP applies whether a party has a full or partial implementation. Parties get the same irrevocable promise from Microsoft either way.

Comments (34)
  1. Simon Phipps says:

    Hi Jason. I’ve been watching the Microsoft patent covenants as they’ve evolved and it’s certainly a much better piece of work than the first attempt – well done. I do however have some comments – apologies if I have missed a development somewhere that addresses them, I admit I have been focusing elsewhere recently.

    * You say "The OSP applies whether a party has a full or partial implementation. Parties get the same irrevocable promise from Microsoft either way." However, OOXML is also covered by CNS which (last time I saw it, I can’t find it now I sit here and type) requires a compatible implementation. Which agreement takes precedence?

    * The OSP says "To clarify, “Microsoft Necessary Claims” are those claims of Microsoft-owned or Microsoft-controlled patents that are necessary to implement only the required portions of the Covered Specification that are described in detail and not merely referenced in such Specification."  As I recall, there are a number of places where OOXML references Microsoft-controlled specifications which are not covered by the OSP. Surely these create an exposure? Would it not be better for this one specification to also extend the OSP to referenced specifications too?

    * As I recall, OOXML contains copyrighted graphics (for border patterns, for example). Since the OSP only refers to patents and not to copyrights, surely a full implementation requires a copyright license from Microsoft?

    I’d be pleased to see these cleared up. Thanks!

  2. hAl says:

    Allthough I think there is no real IPR issue it is Microsofts own fault that the issue is still mentioned.

    It focusses on several points.

    The mentioning/refereral to some MS technologies in the specification (which I noticed has been changed in the current Ecma suggested disposition) and the wording of the OSP which is possibly ambiguous.

    As the interpretation of the OSP as explained by Microsoft can really on be deemed as to make the specifcation fully open it is certain that any legal descision would always take the most favorable interpretation to any implementer there is little problem in using the format specification but the OSP could have been phrased better as was also discusses in a article by Rick Jeliffe recently.

    So even though no real problem exists MS could have been smarter in their attitude to this.

    You still could even, not just by changing the mentioni9ng of MS technology in the format specification but also by either improving on the OSP text itself or on the interpretation of the OSP on the OSP page FAQ.  

    See also:

  3. gerd says:

    Just read what Rex Jaeschke wrote in his ECMA matainance proposal: ECMA IPR policies, i.e. RAND apply.

    Only RF is acceptable and the OSP is no RF patent license. Even when Microsoft played fair how can we be sure there will be no Intellectual Ventures attack on the format etc. etc. Do you trust sharks like Robertson?

    " It is especially convenient for open source software developers as there is no issue as to whether or not the IP is sub-licenseable."

    There is indeed no issues, in fact it is NOT sublicensable, thus useless.

  4. anonymous says:

    You are assuming, as most people are, that any patent issues (presumably copyright and trademark are not issues) would be due to Microsoft patents.

    No-one can state definitively that "there are no patent issues with

    OOXML" since it’s impossible to prove that some unknown 3rd party is not sitting on relevant patents.

    Has anyone made a scenario for what would happen if a random patent holder in country X stepped forwards and claimed that OOXML was infringing?  How would this affect the ISO process?

  5. RichL says:

    OOXML includes other microsoft technologies but doesn’t fully define them in the spec.  Does this IPR protection extend to those as well or only to the OOXML spec proper?  I think the concern related to the incorporated technologies as well as the spec.

  6. Hi Jason,

    what data do you have to backup your claim that "Adoption of OOXML […] is growing exponentially"?

    And I’m not a lawyer either but I know enough about the legal aspect to know that the devil is in the detail. For one thing while it is nice that Microsoft makes "Any Required Microsoft Patent Rights […] Available On A Royalty-Free, Perpetual Basis", as I’m sure you know, the question this raises is what is considered "required".

    There is what is technically required, and there is what is practically required for anyone who wants to compete with Microsoft.

    So you can keep claiming that all this is FUD but they are nonetheless valid concerns people have and ought to have.



  7. hAl says:


    The patent claims should be about implementing or using all of the specified format not about competing with Microsoft.

    I have not heard anyone yet that could formulate any part of the OOXML format that they cannot implement because of the IP rights and thusfar noone implementing the spec has stated such a possible problem.

  8. jasonmatusow says:

    Hi all – good questions.

    1) Anon and Gerd both raise the idea that Microsoft patents may not be the only issue. You guys are right – that is the case with all standards specifications. In all cases, when the parties involved with the production of a specification make their patent declarations, or when the contributor defines the terms covering a spec, they can only speak for themselves and those involved in the process. There is always a risk that others might have claims against the specification. This issue does not constitute an issue for Open XML any more than it would for every other specification out there.

    Simon – you asked about a few things. This answer is about the CNS. My understanding is that it is the choice of the implementor what terms they would like to adhere to. The OSP is the most permissive, and thus the most likely to be used. The fact is, we used the CNS first and we can’t undo its presence in the marketplace. It is not a question of precedence – the OSP is available for anyone to use.

    Arnaud – We have been posting new Open XML implementation solutions all over the world. There are a growing number of solutions every day. I included a good list in the posting. THe fact that there are more than 2 independent implementation efforts is usually a good sign for a spec. In this case there are hundreds. So – pretty good. (Same claim you guys make for ODF if I may point out.)

    As for "required" – I am not a lawyer so I can’t give you a specific answer on this. But, that is COMMON practice for standards IPR policies. Why that is all of a sudden a different issue for Open XML doesn’t make sense.

    At the end of the day, the big question has been, will Microsoft sue me if I go and build an Open XML implementation. The answer is no. And, we have made that statement in an irrevocable, legally binding promise.

    Simon, you asked two other questions – I’m asking my legal beagles for their opinion. I’ll post something back to this thread.



  9. @Simon,

    Maybe ODF implementers also need to get a copyright license from Sun from the derivative schema tags from OpenOffice, for example, the "ole-object-count" tag, but wait, no, that might be from Microsoft too.

    Maybe they should check with Microsoft (or IBM, can’t remember) on using the "dde-source", "dde-connection" and "dde-links" tags as well.

    How about "named-range", the Excel construct?

    There are probably a bunch more if you looked!

    One is always likely to find bizarre edge cases in these sorts of things, less likely to be the result of a PR-suicide lawsuit, I would venture.


  10. RichL,

    "OOXML includes other microsoft technologies but doesn’t fully define them in the spec."

    Could you name a few and cite where in the spec they are located?

    … and could you please be a bit more specific regarding your words "includes other microsoft technologies" – what do you mean with "includes"? You do realise, that OOXML is just a document format, right?


    Jesper Lund Stocholm


  11. jasonmatusow says:

    Simon –

    Again, thank you for the comments. I can always trust you to ask the hardest questions that come in off my blog. Nice to see you’re maintaining your batting average. I think your question about referenced elements should be addressed in two ways. The first is conclusive, the second will merit some longer term consideration.

    First, there is the question of the IPR policies and considerations for national bodies as they think about the merits of DIS 29500. DIS 29500 meets all ISO/IEC JTC1 requirements in relation to intellectual property. If DIS 29500 is to be considered for international standardization by the same guidelines as any other specification, there is no IP-related reason not to vote in favor of its standardization. The whole point of the JTC1 process is to keep deliberations focused on what is in scope and what is out of scope for a given vote or committee meeting or whatever.

    Second, there is a question of the implications for an open source developer who is building DIS 29500 and whether or not referenced specifications have the rights necessary for them to implement under Free Software or Open Source licensing regimes. This is a far more complicated question for a number of reasons. There is much more to be written on this subject, so for the sake of getting you some sort of a response (I know it is a punt of sorts for the moment) I think the most important thing to say is that Microsoft has promised not to sue anyone for the implementation of Open XML – period. Any developer, in any country, implementing Open XML for any application or on any platform – ANY MS patent touched upon during the full or partial implementation of Open XML (“Covered Specification”) may be used with no concern of Microsoft making an IP claim against them. If a reference is made in the specification and is required for implementation of the spec, then we consider it part of the implementation. But, that would be for the implementation of Open XML. There is a wide open question for lawyers to bicker about when you get to the idea of transitive rights where someone makes use of a technology that was referenced in the Open XML spec, but use it for another means, yet say that it was part of an Open XML implementation. This is why I’m not a lawyer. I’ll come back though to the core statement that MS made about Open XML to all developers. You are welcome to build Open XML using any model you so desire.

    Thx – Jason

  12. Doug Mahugh says:

    Apparently the lack of IPR issues around Open XML ( Oliver and Jason have recent coverage of this non-issue)

  13. Apparently the lack of IPR issues around Open XML ( Oliver and Jason have recent coverage of this non

  14. Simon Phipps says:

    Thanks, Jason. Glad to provide a service 🙂  Since the IPR regimes of the standards bodies remain deeply flawed in guaranteeing the freedoms of developers of free/open source software, it’s the second part of your reply that interests me most.

    Will you be adding that guidance to the official site? My experience of community questions about Sun’s policies is that my answers alone don’t satisfy them, they want to see the guidance alongside the covenant. We’ve had some success with the approach of adding an exhaustive Q & A alongside the covenant or license in several contexts and now make that a standard practice.

    You didn’t actually cover my third question (about copyright) – any views on that? Is it an oversight, a misunderstanding or an intentional loophole?

  15. jasonmatusow says:

    Simon, I think there is an important assumption to be clarified in your text. IPR regimes of standards bodies are not designed to guarantee the freedoms of FLOSS developers. They are designed to protect ALL participants – contributors, participants, and implementers alike. That is the whole point of "non-discriminatory." As you are well aware that means there is the potential for direct conflict between those who are patent-holders (who may have a strong affinity for the rights provided to them as patent holders due to their investment in the creation of acquisition of those patents) and those who are anti-patent. I am not seeking to make a value judgement on either side of that, but I think you should be more careful with the description of "deeply flawed."

    As for the copyright question – it was just an oversight on my part. If the spec contains copyrighted graphics then any rights Microsoft had in those copyrights with respect to OOXML were provided to Ecma which in turn has been provided to JTC1 and are available to implementers who obtain their specs from either of those sources.



  16. Simon Phipps says:

    No, I stand by the description. I believe that current IPR policies at many standards bodies encourage the offering of greater freedoms to those operating under closed development practices and proprietary business models than they do to those operating under free/open source software models. They have thus ceased to be "non-discriminatory", although I agree that was the intent in almost all cases.

    I don’t ascribe malice – it’s just they need to catch up with modern practices and add provision for them. I’m not a proponent of discriminating in favour of FOSS, but I do believe that the status quo needs modernisation (for example to encourage sublicensability of rights or equivalent freedoms). Without it, the rules permit discriminatory terms that put FOSS models at a disadvantage.

  17. jasonmatusow says:

    Ok, now we are getting to some common ground. I comletely agree about the need to consider all models and all parties. We have worked very hard, as has Sun, to think about how one bridges the existing constructs of software IP and the new possibilities of collaborative development. The challenge facing any standards org is that any party that does not get what it wants may claim discimination. It is going to take compromise on everyone’s part to find a path that afforts opportunity for all.

    One more important comment about "non-discriminatory." My attorneys have explained to me that it is not about "everyone has access," it is about "everyone has the opportunity for the same access." In otherwords, placing a royalty is not a disciminatory act in and of itself. If I were to offer you royalties at one rate, and the guy next to you a different rate – then that is discriminatory behavior. The fact that a person may not be able to pay that rate is not "discriminatory" in the legal sense of the term. It may create a marketplace adoption blocker though and thus cause the ultimate failure of a spec.

    I do think we need to find a shared path – but it is going to mean that both rights holders and free software advocates alike will have to make concessions.


  18. Simon Phipps says:

    I think you need to consider that a little more, Jason. It’s not about the affordability of the royalty, it is about the uneven application of the requirement for the royalty. The development model behind FOSS does not distinguish between distribution for development and distribution for deployment. However, the payment of royalties does.

    When one works behind a proprietary firewall one pays no royalties to the patent holders of a royalty-bearing standard, and one is (usually) able to pass the source to any other party behind the firewall. However, neither is true of a FOSS development team. To engage in development they have to pay a royalty (since their activity is indistinguishable from "use"), and any person they pass the code to for further work has to as well. Usually, there is not even an entity that could negotiate alternate terms on behalf of the team.

    The only way out of this is to have royalty free, sublicensable standards; without them, FOSS developers do not have "the opportunity for the same access". This is what I mean when I say IP policies need modernisation, since to permit even RAND terms is to exclude FOSS models from use of a standard. Yes, ironically RAND terms are discriminatory even by the definition you cite!

  19. Alors que Jason Matusow publie son post sur la propriété intellectuelle du format Open XML (démontrant

  20. Gerd says:

    Why don’t Microsoft, IBM and SUN get together and sign a joint RF patent protection deal for both formats (DIS 29500 and ISO 26300:2006) that is clearly applicable on a worlwide scale. I think the SUN CNS is fine, the Microsoft CNS looks like a derivate with some trickery, it was abandonned in favour of the other format.

    Is the OSP compatible with EU IDABC rules for "open standards" or not? Many governments are bound by these rules. And I am pretty unsure the OSP is a valid legal document under German law. The estoppel like construction is a concept from anglo-saxon law and it is questionable if you can apply it in continental European law. is a nice collection of existing models for patent indemnification or licensing. It is a frontier of law, so I do not blame Microsoft. We should ensure that no open questions remain and consumer/vendor confidence is provided.

    Hire a set of 100% independent lawyers and let them define a model according to certain goals, e.g. RF, IDABC compliant, defensive counter claims possible, no specification of precise patents, etc etc, business partners are asked to join the agreement to make it more solid plus a legal defence fund for the trolls. Maybe you should ask Barbara Held if the European Commission has resources to develop a standard rf patent model licence.

    Yes, I am very positive that it must be possible to find a fair model for all players, ODF or OOXML aside.

  21. Mark Murphy says:

    "THe fact that there are more than 2 independent implementation efforts is usually a good sign for a spec. In this case there are hundreds"

    You might want to consider building a Web page linking to those "hundreds" of implementations, in particular linking to concrete public evidence of those implementations. Please don’t double-count — I think is included five times in your list, under Novell, NeoOffice (same code base), Xandros/Turbolinux/Linspire (in the form of You only have a dozen or so in your list, which is a far cry from "hundreds".

  22. Jose_X says:

    Looking into the future…

    A standard is useless when one participant owns the majority of the market. At that point all that matters is compatibility with the majority participant. Since Microsoft’s products are closed source, everyone else is effectively unable to provide interoperability within a competitive time frame. It’s clearly in Microsoft’s benefit (and fits well with past behavior) NOT to seek to provide 100% accurate implementation of the standard with no added undocumented features (note that bugs or any ambiguities at all mean less than 100% and any unspecified components of OOXML, if any, would have to be avoided altogether).

    The patent issues are secondary and simply provide another layer of defense for Microsoft since, when they go outside the standard, others will no longer be safe from lawsuits when they try and follow to achieve compatibility. Others might attempt this as they try to reverse-engineer Microsoft "value adds" and "bugs" as well as when they try and keep up with new versions of the standard that may not fall under patent protections.

    These are wonderful games being played over "standards," but all that matters is breaking the monopoly control of a closed source vendor in the various markets that can be (and have been) leveraged to give unfair advantage in related markets. To put it simply, if Microsoft is allowed to compete at making applications and platforms (clients and servers; etc) then Microsoft must be avoided until such time as the monopolies are broken. Openoffice provides a free and open path forward and can understand many existing documents. Alternatively, Microsoft can limit itself to a narrow domain (example, cede the platform advantage, starting by porting Office 2003 to Linux) so that a monopoly there would be less detrimental. Yet another possibility would be to invest in upgrades to Microsoft products only when these products are open source with "no strings attached."

    Truly, this IP issue is quite secondary.

    BTW, what is "Open XML?"

  23. Rick Jelliffe says:

    Trying to penetrate through the legalese and FUD is difficult, even when the OSP tries to be in plain English and the FUD is not always malicious.

    There are some issues that people have raised which are clearly silly. The issue of copyright issue, for example.

    There are other issues that are plausible, but wrong. The idea that the OSP does not cover optional requirement of the normative text, for example.

    However, there are other issues where the OSP does not seem to go far enough. As far as I can see, the OSP (and the equiv from Sun and IBM) only grant patent licenses on technologies that are required to implement the spec: i.e. if there is one way to implement something, and MS etc have patented that way, then the OSP is applicable. But if there are several ways, and MS etc have a patent on one of them, implementers need to pick one of the other ways.

    This is a little different from what Jason is saying here, but it seems to be what the wording of the OSP says. I would much prefer an OSP that makes the coverage of the OSP more explicit,  erring on the wide side.

    Finally, potential users should be aware that ISO does not do an exhaustive independent patent search before allowing a standard. IPR that is known to committee members and editors does need to be disclosed and handled, though.

    In the case of OOXML, document technology and SGML and WP technologies are quite mature: it is hard to imagine many patents being relevant, but it is not impossible. (This applies to HTML 5, PDF, ODF, etc. just as much, and indeed to just making up your own homemade spec too. However, you only expect it would be worth a patent owner’s time to require a license fee when there is some money being made or lost, pre-supposing a successful or mass-market product.)

  24. Dave S. says:

    I was wondering how many are implementations and how many are fragments of implementations.

    I agree with Mark, though – a simple list would be good.

  25. Get A Life! says:

    Are you guys for real?!  Global warming, extreme poverty, hunger, terrosism… and you guys have nothing better to do than argue over document file fomats and the commercial battles between two Goliaths (IBM and MSFT) and one wanna-be (Sun).  Get a life!  Or at least focus your energy on IT industry issue that will have a bigger impact on society – children’s on-line protection, fraud protection, freedom of speech andprivacy in China, etc.

    And yes, what a waste of my time for venturing to read and even posting this.

  26. Orchy Surat says:


    Your proposal is attractive, at least it would reassure the free software guys that there is no problems with patents, because the Software Freedom Law Center gave their green card on the GPL compatibility with Sun promises.

    I would love to see the replacement of the dodgy OSP by a 100% copy of the Sun promises.

    It would give more confidence to people who wants to implement the standard, and shut up the critics (IBM included) of the OSP.

  27. carlos says:

    >These are wonderful games being played over "standards,"

    >but all that matters is breaking the monopoly control of a closed >source vendor in the various markets that can be

    > (and have been) leveraged to give unfair advantage in related


    in government level, this is very difficult. Monopolies generate lot of money $$$ and potliticians love easy money. Open source and competition are useless for them. They perceive it as a menace, no as an opportunity to save people’s tax money.

  28. J-F Bilodeau says:

    Why is there no specific mention of GPL compatibility in the article? This is the point that a lot of OSS developers are debating, and that seems to be the one point you are avoiding. Until that’s resolved, I’m not touching OOXML with a 10-yard stick.

  29. J-F Bilodeau says:

    to Get A Life!,

    Though we may not be directly focusing our energy on children’s on-line protection, fraud protection and privacy in China, I do believe that the fundamental issue is ‘freedom of speech’ a-la GPL.

  30. Alex says:


    I create a docx server and I am trying to sell it, but my customer ask me for an official letter from legal dept in MS that explain in plain english that in the case that they use my server they doesnt have to aqcuaire an Office License or Office Upgrade thinking just in view for print de docx using MS DocViewer. Can you help me with an explanation of what will be the procedure to obtain that kind of letter? or please refer me to an appropiate lawyer firm for this matter.

    Thanks in advance,


Comments are closed.

Skip to main content