ORDER AND MEMORANDUM OPINION RE ATTORNEY-CLIENT PRIVILEGE AND WORK-PRODUCT PROTECTION
BRAZIL, United States Magistrate Judge.
I. FACTUAL BACKGROUND
The discovery disputes addressed in this Opinion and Order arise against a background that includes the following pertinent facts.
The defendant, General Scanning, Inc. (hereafter referred to as GSI), has elected to use two kinds of outside counsel in connection with matters that are relevant to this litigation. GSI refers to one of these kinds of lawyers as its “opinion counsel” — who is John (Jack) N. Williams of the law firm of Fish & Richardson (Boston office). At least six months before plaintiff, Electro Scientific Industries, Inc. (hereafter referred to as ESI) filed this lawsuit,1 GSI retained Mr. Williams to formulate opinions and offer legal advice to GSI about the ESI patents in suit and the relationship between them and GSI’s products/systems. About five months before the complaint was filed, Mr. Williams [540]*540prepared a written “opinion” that he delivered in letter form to GSI. It appears that Mr. Williams later prepared another written opinion, which he had delivered to GSI in early January of this year, less than two weeks after plaintiff filed the complaint that commenced this litigation. How much oral communication, if any, there was between Mr. Williams and GSI on these subjects is unclear.
GSI apparently never retained Mr. Williams to serve as its litigation counsel in connection with this lawsuit. Instead, GSI at some point separately retained the law firm of Brooks & Kushman (of Southfield, Michigan) to perform that function. How much communication there has been between Mr. Williams and the Brooks & Kushman firm is unclear — as is when that communication began — but during the hearing on plaintiffs motion to compel Mr. Brooks implied that at some juncture lawyers from his firm and Mr. Williams communicated substantially.
About two weeks ago GSI, through its litigation counsel (Mr. Brooks), served a “Notice of Intent to Rely on Opinion of Counsel.” In this Notice, GSI declared that it “intends to rely on its July 12, 1996 opinion of counsel letter [from Mr. Williams] to rebut ESI’s allegations of willful infringement.” In this same Notice, GSI went on to announce that it would “produce the July 12, 1996 letter, and all attorney-client communications up to the date this action was filed concerning the subject matter of the opinions expressed therein....” I infer from the language in this Notice that GSI has agreed to disclose in response to discovery requests (for documents and in deposition questions) all communications, written or oral, between GSI, on the one hand, and, on the other, either its “opinion counsel” or its “litigation counsel” — ■ as long as those communications ‘concern’ the subject matter of the July 12, 1996 opinion letter and as long as they occurred before December 26, 1996, the date ESI filed the complaint. This commitment by GSI is driven by its acknowledgment that when it elected to use a pre-litigation opinion of counsel as part of its defense against the charge of willful infringement it put in issue what it had learned from counsel about the matters covered in the July 12th letter, whether the source of what it (GSI) had learned on these subjects was a lawyer denominated “opinion counsel” or a lawyer denominated “litigation counsel.”
It is against this general background that several disputes have arisen. First, ESI apparently contends that the waiver-effected by GSI’s decision to rely on the pre-litigation opinion of counsel is not limited to the period that precedes the filing of the complaint. Second, ESI contends that, wholly apart from GSI’s decision to rely on the opinion, letter of July 12th, there is an independent basis for a finding that GSI has waived the protections of the attorney-client privilege on the substance of its attorney’s opinions. According to ESI, GSI voluntarily disclosed (on several different occasions) the contents of otherwise privileged communications from its opinion counsel on these subjects — and thereby must be deemed to have waived its privilege for communications in the same subject areas. According to ESI, the period covered by the waiver based on this independent ground also is not limited by the filing of the complaint. GSI takes the position that the only basis for a finding of waiver is its decision to rely on the July 12th opinion letter' — that none of its other disclosures were sufficient to support a finding of waiver. GSI also contends that most of the disclosures it made were immunized by the “community of interest” doctrine. See, Defendant GSI’s Response to Plaintiff ESI’s Motion to Compel Production, filed August 20, 1997 at 4:3 — 5:2; see also Paul R. Rice, Attorney-Client Privilege in the United States, (1993) at §§ 4:35 and 4:36 (hereinafter “Rice”); and Hewlett-Packard Co. v. Bausch & Lamb, Inc., 115 F.R.D. 308 (N.D.Cal.1987). In addition, GSI argues that any waiver that is found to have been caused by any of its earlier disclosures must be limited to the period before the complaint was filed or, if made thereafter, to the period before GSI made the specific disclosure that would serve as the basis for the finding of the waiver.
The other dispute raised by the parties’ papers relates to the work product doctrine. While conceding that, because it has decided to rely on the July 12th opinion from its [541]*541counsel, it must disclose “communications” that would otherwise be protected by either the attorney-client privilege or the work product doctrine, GSI contends that papers generated in anticipation of litigation by either its “opinion counsel” or its “litigation counsel” that do not reflect “communications” with GSI remain insulated from discovery by the work product doctrine. ESI disagrees.
In this Opinion and Order I resolve the disputes about (1) whether disclosures independent of the decision to rely on the July 12th opinion letter were sufficient to support a finding of waiver of the attorney-client privilege and (2) whether the decision to rely on advice of counsel requires limited disclosure of documents that might otherwise be protected by the work product doctrine. With one important exception,2 I decline to resolve the parties’ disputes about the temporal reach of the waivers I find. For the most part, the “temporal reach” issues were not the subject of briefing, even though they appear to be both important and complicated. I will not reach these issues unless and until the parties develop the kind of record (factual and legal) that could support a more reliable analysis by the court.
II. WAIVER OF ATTORNEY-CLIENT PRIVILEGE BY DISCLOSURE OF THE CONTENT OF OTHERWISE PRIVILEGED COMMUNICATIONS
As noted above, ESI contends that, independent of GSI’s decision to rely on the July 12th opinion of counsel, on several occasions GSI voluntarily made disclosures of otherwise privileged communications from counsel and thereby waived the protections of the attorney-client privilege for communications about subjects addressed in the disclosed communications. In response, GSI relies primarily on two arguments: (1) GSI contends that the disclosures it made did not reveal the substance of any communications from counsel — or at least that not much of the substance was disclosed and certainly no more than could be readily inferred from the answer that GSI filed in this action; (2) GSI contends that any waiver that might otherwise be found did not occur because the disclosures made by GSI (even if substantive) were protected under the “community of interest” doctrine.
ESI has identified several disclosures by GSI (apart from the decision to rely on the opinion letter of July 12, 1996) that it contends effect a waiver of GSI’s privilege. These include oral (and perhaps written) communications between opinion counsel for GSI and lawyers for Mitsubishi and for IBM (customers or prospective customers for the accused GSI products/systems), as well as the distribution in late January of 1997 of a “white paper” to representatives of IBM and to representatives of Siemens (both possible buyers of the accused GSI products/systems). It is not clear, on the record before me, when (after July 30, 1996) the oral communications between GSI’s opinion counsel and Mitsubishi and IBM occurred, or what those communications included. In sharp contrast, it is clear what was in the “white paper” and when it was distributed. The parties have submitted to the court two versions of the “white paper,” one that GSI presented to IBM on January 29, 1997, and a second, modestly edited version that GSI presented to Siemens on January 30, 1997. In these white papers, GSI voluntarily (for the purpose of advancing its own commercial interests) disclosed substantive components of what otherwise would have been privileged communications from its opinion counsel.3 [542]*542Those disclosures, however, were made under circumstances (including steps taken to maintain confidentiality) that probably would support a finding that the “community of interest” doctrine applied — -with the result that no waiver would be effected by these particular communications. The community of interest doctrine also probably would reach the oral communications between Mr. Williams, on the one hand, and the legal representatives of Mitsubishi and IBM, on the other.
There is an additional disclosure by GSI, however, to which the “community of interest” doctrine does not even arguably apply. On January 7, 1997, GSI issued a “News Release” entitled “General Scanning Responds to Patent Infringement Claims.” In this document, GSI declared that it “has previously carefully studied ESI’s patent claims and has been advised by legal counsel that the referenced patents are invalid based upon prior patents and publications not cited to the Patent Office.” See, Ex. A to Declaration of William L. Leschensky in Support of Reply Memorandum of Plaintiff ESI in Support of Motion to Compel Production of Documents from Defendant GSI and Request for Sanctions, filed August 27,1997. There is no evidence that this “News Release” was disclosed in strict confidence to only a limited number of GSI customers — so GSI cannot [543]*543plausibly contend that the “community of interest” doctrine applies to it.
What the “News Release” disclosed about the otherwise confidential opinion of counsel clearly was substantive — -and more specific than anything in the Answer that GSI filed two weeks later. That Answer asserted generally that the ESI patents were unenforceable because ESI had engaged in unspecified “inequitable conduct before the U.S. Patent Office” and “for failure to meet one or more of the conditions for patentability specified in Part II of Title 35 of the United States Code and/or for failure to comply with one or more of the requirements of 35 U.S.C. § 112.” But the “News Release” disclosed that GSI’s attorney had expressed an opinion to GSI (in what would have been a confidential communication) that the reason the patents were not enforceable was that ESI had failed to cite “prior patents and publications” to the patent examiner.
I hasten to emphasize, moreover, that the fact that a party takes a position in a pleading that is consistent with advice that party received in confidence from its attorney is irrelevant to waiver analysis. Waiver analysis focuses on the disclosure of the content of specific communications between counsel and client — and a pleading would not effect a waiver unless the pleading disclosed specific lawyer-client communications, even if the substance of the pleading tracked what a lawyer had confidentially advised the client.
The fact that is critical to waiver analysis in the situation at bar is the fact that the “News Release” voluntarily disclosed an important, substantive component of a communication from counsel. What the “News Release” disclosed from that communication was the most important part of it: the bottom line of the lawyer’s opinion, his conclusion, the ultimate outcome of his legal reasoning. The “News Release” disclosed that the lawyer had advised the client, in presumptive confidence, that, from among the many theoretically available grounds for defending the lawsuit, the lawyer’s advice was that the defense should focus on the theory that ESI had failed to bring pertinent prior art (patents and publications) to the attention of the examiner.
These are substantive disclosures of counsel’s advice — and the absence of the details of the reasoning that accompanied counsel’s advice does not compel the conclusion that not enough substance was disclosed to support a finding of waiver. I know of no precedents that hold that all of a communication between attorney and client must be disclosed before a waiver occurs, and it makes no sense to hold that no waiver occurs when what is disclosed is the most important part of the privileged communication, but not the details. A sophisticated, well-counseled party who intentionally discloses an important part of an otherwise privileged communication acts in a manner that is .thoroughly inconsistent with preserving the confidentiality of that communication. Stated somewhat differently, a sophisticated party who intentionally discloses the most significant part of an otherwise privileged communication, in an act calculated to advance that party’s commercial interests, cannot establish, as the law would require, that the party reasonably believed that it would be able to preserve the confidentiality of the other parts of that communication. For these reasons, I hold that the “News Release” effected a waiver of attorney-client privilege.
I next must determine how broad that waiver is, i.e., which communications it reaches. This is an issue that the law commits in some measure to my discretion. See, Rice at §§ 9:30, 9:77, and 9:78. In exercising that discretion, considerations of fairness should play the most significant role. At this juncture in the proceedings, as I understand it, GSI has not elected to use for any purpose in this litigation the sentence in the “News Release” that discloses the communication from counsel. As long as that remains true, the disclosure of the communication would remain “extrajudicial” — and would pose no risk of “truth garbling” in the administration of justice. See, Rice at §§ 9:28 and 9:80; see also Richard Marcus, The Perils of Privilege: Waiver and the Litigator, 84 Mich. L.Rev. 1605 (1986). And while the disclosure in the “News Release” certainly appears to have been calculated (rather than inadvertent), and obviously was intended to advance GSI’s commercial interests, I cannot see how that [544]*544past disclosure, standing alone, would cause any prejudice in this litigation to ESI. It follows that as long as GSI does not try to use in this litigation the sentence in the “News Release” that discloses the communication from counsel, the scope of the waiver should be narrow. At this juncture, GSI has waived its privilege with respect to all of the communication or communications from counsel to which it adverted or on which it relied when it declared in its “News Release” of January 7, 1997, that “it has been advised by legal counsel that the referenced patents are invalid____”
If GSI elects to use that sentence from the “News Release” in this litigation, the disclosure would become “judicial” (would no longer remain “extrajudicial”) and would create a serious risk of truth garbling and prejudice to ESI. In that event, the scope of the waiver would be broader and would reach, at least, all communications about any of the subjects that were addressed in the communication or communications that underlay GSI’s statement that “it has been advised by legal counsel that the referenced patents are invalid....”
III. EFFECT OF RELIANCE ON ADVICE OF COUNSEL ON WORK PRODUCT PROTECTIONS
GSI has elected to use a July 12, 1996 letter from its “opinion counsel” as part of its defense against ESI’s allegation of willful infringement. GSI concedes that by electing to use this otherwise privileged communication as part of its defense, GSI has waived the protections of the attorney-client privilege with respect to all communications with counsel that precede the filing of the complaint in this action and that relate to subjects addressed in the July 12th opinion letter. GSI contends, however, that its election to use the July 12th letter leaves fully intact the protections that it or its counsel enjoy under the work product doctrine. ESI counters by arguing that when GSI decided to rely on the opinion letter, GSI also waived its protections under the work product doctrine.
There is a division of opinion in the authorities on this matter. At least three courts, in what seem to be carefully reasoned opinions, have concluded that work product materials need not be produced in a setting like this. See, Steelcase Inc. v. Haworth, Inc., 954 F.Supp. 1195 (W.D.Mich.1997); Thorn EMI North America, Inc. v. Micron Technology, 837 F.Supp. 616 (D.Del.1993); Micron Separations, Inc. v. Pall Corporation, 159 F.R.D. 361 (D.Ma.1995). In essence, these courts reason as follows: when litigating willful infringement, the mind that is in issue is the mind of the allegedly infringing party, not the separate mind of that party’s lawyer; in order to explore the mind of the alleged infringer, it is necessary to know all the information that reached that mind (including attorney-client communications, when advice of counsel is part of the defense), but information that did not reach that mind is irrelevant and, therefore, not discoverable. Under this line of reasoning, documents in the lawyer’s files that constitute work product but that do not reflect communications with the client remain protected from discovery.
A considerable number of opinions come to a contrary conclusion, holding generally that when a party elects to rely on advice of counsel that party waives its protections under both the attorney-client privilege and the work product doctrine. E.g., Mushroom Assoc. v. Monterey Mushrooms, Inc., 24 U.S.P.Q.2d 1767 (N.D.Cal.1992); Haney v. Timesavers, Inc., 1995 WL 604100 (D.Ore. 1995); Matsushita Electronics Corporation v. Loral Corporation, 1995 WL 527640 (S.D.N.Y. 1995). These opinions, however, do not squarely address the “irrelevance” rationale described in the preceding paragraph.
For reasons I will explain, I have concluded that GSI must disclose some documents that would otherwise be protected by the work product doctrine. But that disclosure will be subject to an “attorney’s eyes only” protective order, meaning that only outside counsel for ESI will have access to these documents unless otherwise ordered or stipulated. And the court will issue an order permitting ESI’s counsel to use any of this material in this litigation only on a showing that it has potentially significant probative value, as explained below.
[545]*545The reasoning that supports my disposition of this matter- is as follows. At the outset, I agree fully that in litigating the willful infringement issue, it is GSI’s mind, not the mind of any of its lawyers, that is in issue, i.e., that is the ultimate target of inquiry. But I do not believe it follows from that proposition that documents in counsel’s files that do not directly disclose communications that reached GSI could not have any significant probative value.
GSI is claiming that it relied in part on the opinion of its counsel that is set forth in the letter of July 12, 1996. In taking that position, GSI at least implicitly also is asserting that it received no other (oral or written) opinions from counsel, at least before the lawsuit was filed, that were in any material respects different from the opinions reflected in the July 12th letter. In other words, GSI is claiming that this one letter reflects accurately the substance of all of the communications from counsel that reached GSI’s mind (before the suit was filed) on whether GSI would be violating ESI’s rights under ESI’s patents if GSI continued to attempt to manufacture, use, or market products or systems covered by ESI’s patents.
How is the trier of fact to know whether this implicit assertion by GSI is true? Is it fair, in circumstances like this, to cut-off ESI’s access to evidence that might well have significant probative weight about whether this important representation by GSI is accurate? Documents in the files of GSI’s counsel might be highly probative of whether GSI in fact received additional communications from counsel about the issues addressed in the opinion letter, and about whether any such additional communications included views, information, or analysis that was in any material way different from the views, information or analysis in that letter.
GSI argues that documents in its outside4 lawyers’ (whether “opinion counsel” or “litigation counsel”) files that do not directly reflect or advert to communications that reached GSI could not shed any significant light on the issue of whether GSI received additional communications and, if so, whether their content was meaningfully different from the content of the one disclosed communication. I disagree — not as a matter of likelihood, but as a matter of real possibility. It is possible that documents in opinion counsel’s work product file could reflect very different analyses and conclusions than were set forth in the one disclosed letter. The fact that the analyses and conclusions in the lawyer’s private file were clearly at odds with the content of the disclosed opinion would tend to support an inference that there were additional communications between client and counsel and that in those communications the client received opinions that were not consistent with the views expressed in the disclosed letter.
Certainly it would not be rational to assume that everything in counsel’s files reached the client, or that counsel communicated to the client all of what he or she really thought, but it would be comparably irrational to assume that there could be no relationship between what counsel really thought (as reflected in her private papers) and what she in fact communicated to her client. In this important sense, evidence about what really was in the lawyer’s mind could be quite relevant to the issue of what really was in the client’s mind.
Moreover, it is impossible to foresee all the kinds of evidence that might be found in the kind of work product documents that are in issue here (ie., documents that do not directly reflect or advert5 to communications with [546]*546the client), or all the kinds of ways that the contents of such documents might be used to shed reliable light on the issue of whether the defendant received additional opinions (perhaps orally) that were inconsistent with the opinions reflected in the disclosed opinion letter. So, because we can foresee at least one way in which such private documents might be highly probative on the issue of what really reached the mind of the client, and because it certainly is possible that there are other ways such documents might contribute in important ways to determining the truth about the state of mind of the client, it would be unfair, at least absent a compelling countervailing interest, to cut-off access to evidence and leads of this kind.
A conclusion that such material might well be relevant, or might well lead to the discovery of admissible evidence, does not, of course, end our inquiry. We still must ask whether it is appropriate to penetrate, at least to some extent, the work product protection. In answering this question, we appreciate that different levels of protection attach to different kinds of work product (a much stronger showing is necessary to discover “opinion” work product than to discover “non-opinion” work product). To expedite the discussion, however, I will assume that the kind of work product most likely to be in issue in settings like this is “opinion” work product — the kind that enjoys the highest level of protection. Working from this assumption, we frame the issue as follows: are the values and interests that would be promoted or protected by ordering limited access to opinion work product in this situation appreciably weightier than the harm that would be done to the values that support protection for opinion work product?
I conclude that in circumstances like those presented here, the answer to this question is yes — at least when the court orders, as I do here, that the work product will be disclosed (absent further order of the court or stipulation) only to plaintiffs outside counsel. The issue to which the work product materials would relate is extremely important: a finding of willfulness exposes an infringer to treble damages, and in many patent cases very large sums of money are at stake, to say nothing of the reputations of both persons and companies. And, for reasons set forth above, it is possible that the kind of work product evidence that is in issue here could, in some cases, contribute a great deal to the reliability of the truth-finding process with respect to this important issue.
On the other hand, it is not clear that a limited disclosure order like the one I will enter here would do serious harm to the values that support the work product doctrine. Even if they knew that there was some risk that their pre-lawsuit work would be exposed if their client were to elect to rely on opinion of counsel, lawyers who develop advice for clients about infringement, validity and enforceability would continue to have substantial incentives to conduct thorough and thoughtful analyses. Counsel will continue to want their analyses to be reliable— in part as a matter of professional pride and service, and in part to avoid malpractice actions. So the possibility of exposure of prelitigation work product on these issues should not do serious harm to the incentives that have been considered important to the vitality of the adversary system.
Nor does it seem likely that ordering limited disclosure of the kind of work product in issue here (documents that reflect or relate to pre-litigation analyses of infringement, validity and/or enforceability) is likely to jeopardize the truth finding process by enabling one side to learn the details of the other side’s trial tactics. It could be argued that opponents who were able to learn such details would be able to prepare cynical parries to the questions on cross-examination that would otherwise expose untruthful witnesses. My order, however, will only reach documents that were prepared before the lawsuit was filed and that relate to the analyses of infringement, validity, and/or enforceability — so documents with more direct tactical sensitivity, like outlines of possible questions for deponents or trial witnesses, or compilations of materials to be used to support an argument to the jury, would remain off limits.
[547]*547There is a greater risk that the kind order I am entering here will cause harm to the interests in creativity and candor that are advanced when lawyers believe that their analyses and conclusions will remain private. Again, however, it is important to bear in mind that this order will result in disclosure of no work product materials that were generated after the lawsuit was filed. Of course, our system also has an interest in encouraging lawyers to brainstorm, ruminate, and undertake frank analyses before lawsuits are filed. And my disposition of this part of plaintiffs motion will harm that interest. How much harm it will do I am not sure, but, as explained in the next paragraph, the order I am fashioning will limit that harm except in the probably unusual case where the contribution to the truth finding process that would be made by permitting litigation use of the work product materials will visibly outweigh the competing work product interests. of
Under the order I am entering today, only ESI’s outside litigation counsel will have access to this work product. Neither ESI nor anyone else will see this material unless and until GSI agrees to wider dissemination or the court, after additional adversary proceedings, expressly authorizes additional use of some of the material. If the work product materials that are disclosed by GSI’s opinion counsel generally support GSI’s position— that it received no opinions (before the lawsuit was filed) that were materially inconsistent with the July 12th opinion letter, then the work product materials will never see the light of day. The court will permit ESI’s outside counsel to use the work product materials in this litigation only if those materials lend meaningful support to an inference that GSI in fact received opinions or information that are materially different from the opinions or information in the one disclosed opinion letter. Thus, it is only when they would contribute significantly to the reliability of the truth finding process with respect to an important issue that the work product materials from GSI’s counsel would be disclosed to anyone other than ESI’s outside counsel— and in that setting, the values promoted by the disclosure clearly would outweigh the values harmed by it.
For the reasons set forth above, I ORDER counsel (whether “opinion counsel” or “litigation counsel”) for GSI to disclose only to outside counsel for ESI materials that (1) were generated before this lawsuit was filed, and (2) that would otherwise be protected by the work product doctrine, and (3) that reflect or relate to matters addressed in the July 12th opinion letter. ESI’s outside counsel may disclose the contents of any of the material produced in response to this order only on stipulation by GSI or further order of the court.
SANCTIONS
In clear violation of Civil Local Rule 37-1(e), counsel for ESI added a sentence to the end of its Reply Brief asking for cost and fee shifting sanctions. That request is DENIED, with an admonition that if a request for sanctions is similarly made in the future sanctions will be imposed on counsel for ESI.
IT IS SO ORDERED.