Opinion for the court filed by Circuit Judge MAYER, in which Circuit Judges NEWMAN, LOURIE, RADER, SCHALL, BRYSON, GAJARSA, LINN, DYK, and PROST join. Concurring opinion filed by Circuit Judge GAJARSA, in which Circuit Judge NEWMAN joins. Concurring opinion filed by Circuit Judge NEWMAN.
MAYER, Circuit Judge.
Seagate Technology, LLC (“Seagate”) petitions for a writ of mandamus directing the United States District Court for the Southern District of New York to vacate its orders compelling disclosure of materials and testimony that Seagate claims is covered by the attorney-client privilege and work product protection. We ordered en banc review, and now grant the petition. We overrule Underwater Devices Inc. v. Morrison-Knudsen Co., 717 F.2d 1380 (1983), and we clarify the scope of the waiver of attorney-client privilege and work product protection that results when an accused patent infringer asserts an advice of counsel defense to a charge of willful infringement.
[1366]
Background
Convolve, Inc. and the Massachusetts Institute of Technology (collectively “Convolve”) sued Seagate on July 13, 2000, alleging infringement of U.S. Patent Nos. 4,916,635 (“the '635 patent”) and 5,638,267 (“the '267 patent”). Subsequently, U.S. Patent No. 6,314,473 (“the '473 patent”) issued on November 6, 2001, and Convolve amended its complaint on January 25, 2002, to assert infringement of the '473 patent. Convolve also alleged that Sea-gate willfully infringed the patents.
Prior to the lawsuit, Seagate retained Gerald Sekimura to provide an opinion concerning Convolve’s patents, and he ultimately prepared three written opinions. Seagate received the first opinion on July 24, 2000, shortly after the complaint was filed. This opinion analyzed the '635 and '267 patents and concluded that many claims were invalid and that Seagate’s products did not infringe. The opinion also considered Convolve’s pending International Application WO 99/45535 (“the '535 application”), which recited technology similar to that disclosed in the yet-to-be-issued '473 patent. On December 29, 2000, Sekimura provided an updated opinion to Seagate. In addition to his previous conclusions, this opinion concluded that the '267 patent was possibly unenforceable. Both opinions noted that not all of the patent claims had been reviewed, and that the '535 application required further analysis, which Sekimura recommended postponing until a U.S. patent issued. On February 21, 2003, Seagate received a third opinion concerning the validity and infringement of the by-then-issued '473 patent. There is no dispute that Seagate’s opinion counsel operated separately and independently of trial counsel at all times.
In early 2003, pursuant to the trial court’s scheduling order, Seagate notified Convolve of its intent to rely on Sekimu-ra’s three opinion letters in defending against willful infringement, and it disclosed all of his work product and made him available for deposition. Convolve then moved to compel discovery of any communications and work product of Sea-gate’s other counsel, including its trial counsel.1 On May 28, 2004, the trial court concluded that Seagate waived the attorney-client privilege for all communications between it and any counsel, including its trial attorneys and in-house counsel,2 concerning the subject matter of Sekimura’s [1367] opinions, i.e., infringement, invalidity, and enforceability. It further determined that the waiver began when Seagate first gained knowledge of the patents and would last until the alleged infringement ceased. Accordingly, the court ordered production of any requested documents and testimony concerning the subject matter of Sekimu-ra’s opinions. It provided for in camera review of documents relating to trial strategy, but said that any advice from trial counsel that undermined the reasonableness of relying on Sekimura’s opinions would warrant disclosure. The court also determined that protection of work product communicated to Seagate was waived.
Based on these rulings, Convolve sought production of trial counsel opinions relating to infringement, invalidity, and enforceability of the patents, and also noticed depositions of Seagate’s trial counsel. After the trial court denied Seagate’s motion for a stay and certification of an interlocutory appeal, Seagate petitioned for a writ of mandamus. We stayed the discovery orders and, recognizing the functional relationship between our willfulness jurisprudence and the practical dilemmas faced in the areas of attorney-client privilege and work product protection, sua sponte ordered en banc review of the petition. The en banc order set out the following questions:
1. Should a party’s assertion of the advice of counsel defense to willful infringement extend waiver of the attorney-client privilege to communications with that party’s trial counsel? See In re EchoStar Commc’ns Corp., 448 F.3d 1294 (Fed.Cir.2006).
2. What is the effect of any such waiver on work-product immunity?
3. Given the impact of the statutory duty of care standard announced in Underwater Devices, Inc. v. Morrison-Knudsen Co., 717 F.2d 1380 (Fed.Cir.1983), on the issue of waiver of attorney-client privilege, should this court reconsider the decision in Underwater Devices and the duty of care standard itself?
In re Seagate Tech., LLC, 214 Fed.Appx. 997 (Fed.Cir.2007).
Mandamus
A party seeking a writ of mandamus bears the burden of proving that it has no other means of attaining the relief desired, Mallard v. U.S. Dist. Court for the S. Dist. of Iowa, 490 U.S. 296, 309, 109 S.Ct. 1814, 104 L.Ed.2d 318 (1989), and that the right to issuance of the writ is “clear and indisputable,” Allied Chem. Corp. v. Daiflon, Inc., 449 U.S. 33, 35, 101 S.Ct. 188, 66 L.Ed.2d 193 (1980). In appropriate cases, a writ of mandamus may issue “to prevent the wrongful exposure of privileged communications.” In re Regents of the Univ. of Cal., 101 F.3d 1386, 1387 (Fed.Cir.1996). Specifically, “mandamus review may be granted of discovery orders that turn on claims of privilege when (1) there is raised an important issue of first impression, (2) the privilege would be lost if review were denied until final judgment, and (3) immediate resolution would avoid the development of doctrine that would undermine the privilege.” Id. at 1388. This case meets these criteria.
We review the trial court’s determination of the scope of waiver for an abuse of discretion. In re EchoStar Commc’ns. Corp., 448 F.3d 1294, 1300 (Fed.Cir.2006); In re Pioneer Hi-Bred Int’l, Inc., 238 F.3d 1370, 1373 n. 2 (Fed.Cir.2001). Because willful infringement and the scope of waiver accompanying the [1368] advice of counsel defense invoke substantive patent law, we apply the law of this circuit. EchoStar, 448 F.3d at 1298.
Discussion
Because patent infringement is a strict liability offense, the nature of the offense is only relevant in determining whether enhanced damages are warranted. Although a trial court’s discretion in awarding enhanced damages has a long lineage in patent law,3 the current statute, similar to its predecessors, is devoid of any standard for awarding them.4 Absent a statutory guide, we have held that an award of enhanced damages requires a showing of willful infringement. Beatrice Foods Co. v. New England Printing & Lithographing Co., 923 F.2d 1576, 1578 (Fed.Cir.1991); see also Jurgens v. CBK, Ltd., 80 F.3d 1566, 1570 (Fed.Cir.1996) (holding that bad faith infringement, which is a type of willful infringement, is required for enhanced damages). This well-established standard accords with Supreme Court precedent. See Aro Mfg. Co. v. Convertible Top Replacement Co., 377 U.S. 476, 508, 84 S.Ct. 1526, 12 L.Ed.2d 457 (1964) (enhanced damages were available for willful or bad faith infringement); see also Dowling v. United States, 473 U.S. 207, 227 n. 19, 105 S.Ct. 3127, 87 L.Ed.2d 152 (1985) (enhanced damages are available for “willful infringement”); Seymour v. McCormick, 57 U.S. 480, 489, 16 How. 480, 14 L.Ed. 1024 (1853) (“wanton or malicious” injury could result in exemplary damages). But, a finding of willfulness does not require an award of enhanced damages; it merely permits it. See 35 U.S.C. § 284; Odetics, Inc. v. Storage Tech. Corp., 185 F.3d 1259, 1274 (Fed.Cir.1999); Jurgens, 80 F.3d at 1570.
This court fashioned a standard for evaluating willful infringement in Underwater Devices Inc. v. Morrison-Knudsen Co., 717 F.2d 1380, 1389-90 (Fed.Cir.1983): “Where ... a potential infringer has actual notice of another’s patent rights, he has an affirmative duty to exercise due care to determine whether or not he is infringing. Such an affirmative duty includes, inter alia, the duty to seek and obtain competent legal advice from counsel before the [1369] initiation of any possible infringing activity.” (citations omitted). This standard was announced shortly after the creation of the court, and at a time “when widespread disregard of patent rights was undermining the national innovation incentive.” Knorr-Bremse Systeme Fuer Nutzfahrzeuge GmbH v. Dana Corp., 383 F.3d 1337, 1343 (Fed.Cir.2004) (en banc) (citing Advisory Committee on Industrial Innovation Final Report, Dep’t of Commerce (Sep.1979)). Indeed, in Underwater Devices, an attorney had advised the infringer that “[c]ourts, in recent years, have — in patent infringement cases — found [asserted patents] invalid in approximately 80% of the cases,” and on that basis the attorney concluded that the patentee would not likely sue for infringement. 717 F.2d at 1385. Over time, our cases evolved to evaluate willfulness and its duty of due care under the totality of the circumstances, and we enumerated factors informing the inquiry. E.g., Read Corp. v. Portec, Inc., 970 F.2d 816, 826-27 (Fed.Cir.1992); Rolls-Royce Ltd. v. GTE Valeron Corp., 800 F.2d 1101, 1110 (Fed.Cir.1986).
In light of the duty of due care, accused willful infringers commonly assert an advice of counsel defense. Under this defense, an accused willful infringer aims to establish that due to reasonable reliance on advice from counsel, its continued accused activities were done in good faith. Typically, counsel’s opinion concludes that the patent is invalid, unenforceable, and/or not infringed. Although an infringer’s reliance on favorable advice of counsel, or conversely his failure to proffer any favorable advice, is not dispositive of the willfulness inquiry, it is crucial to the analysis. E.g., Electro Med. Sys., S.A. v. Cooper Life Scis., Inc., 34 F.3d 1048, 1056 (Fed.Cir.1994) (“Possession of a favorable opinion of counsel is not essential to avoid a willfulness determination; it is only one factor to be considered, albeit an important one.”).
Since Underwater Devices, we have recognized the practical concerns stemming from our willfulness doctrine, particularly as related to the attorney-client privilege and work product doctrine. For instance, Quantum Corp. v. Tandon Corp., 940 F.2d 642, 643 (Fed.Cir.1991), observed that “[p]roper resolution of the dilemma of an accused infringer who must choose between the lawful assertion of the attorney-client privilege and avoidance of a willfulness finding if infringement is found, is of great importance not only to the parties but to the fundamental values sought to be preserved by the attorney-client privilege.” We cautioned there that an accused in-fringer “should not, without the trial court’s careful consideration, be forced to choose between waiving the privilege in order to protect itself from a willfulness finding, in which case it may risk prejudicing itself on the question of liability, and maintaining the privilege, in which case it may risk being found to be a willful in-fringer if liability is found.” Id. at 643-44. We. advised that in camera review and bifurcating trials in appropriate cases would alleviate these concerns. Id. However, such procedures are often considered too onerous to be regularly employed.
Recently, in Knorr-Bremse, we addressed another outgrowth of our willfulness doctrine. Over the years, we had held that an accused infringer’s failure to produce advice from counsel “would warrant the conclusion that it either obtained no advice of counsel or did so and was advised that its [activities] would be an infringement of valid U.S. Patents.” Knorr-Bremse, 383 F.3d at 1343 (quoting [1370] Kloster Speedsteel AB v. Crucible Inc., 793 F.2d 1565, 1580 (Fed.Cir.1986)). Recognizing that this inference imposed “inappropriate burdens on the attorney-client relationship,” id., we held that invoking the attorney-client privilege or work product protection does not give rise to an adverse inference, id. at 1344^45. We further held that an accused infringer’s failure to obtain legal advice does not give rise to an adverse inference with respect to willfulness. Id. at 1345-46.
More recently, in EchoStar we addressed the scope of waiver resulting from the advice of counsel defense. First, we concluded that relying on in-house counsel’s advice to refute a charge of willfulness triggers waiver of the attorney-client privilege. EchoStar, 448 F.3d at 1299. Second, we held that asserting the advice of counsel defense waives work product protection and the attorney-client privilege for all communications on the same subject matter, as well as any documents memorializing attorney-client communications. Id. at 1299, 1302-03. However, we held that waiver did not extend to work product that was not communicated to an accused infringer. Id. at 1303-04. EchoStar did not consider waiver of the advice of counsel defense as it relates to trial counsel.
In this case, we confront the willfulness scheme and its functional relationship to the attorney-client privilege and work product protection. In light of Supreme Court opinions since Underwater Devices and the practical concerns facing litigants under the current regime, we take this opportunity to revisit our willfulness doctrine and to address whether waiver resulting from advice of counsel and work product defenses extend to trial counsel. See Knorr-Bremse, 383 F.3d at 1343-44.
I. Willful Infringement
The term willful is not unique to patent law, and it has a well-established meaning in the civil context. For instance, our sister circuits have employed a recklessness standard for enhancing statutory damages for copyright infringement. Under the Copyright Act, a copyright owner can elect to receive statutory damages, and trial courts have discretion to enhance the damages, up to a statutory maximum, for willful infringement. 17 U.S.C. § 504(c). Although the statute does not define willful, it has consistently been defined as including reckless behavior. See, e.g., Yurman Design, Inc. v. PAJ, Inc., 262 F.3d 101, 112 (2d Cir.2001) (“Willfulness in [the context of statutory damages for copyright infringement] means that the defendant ‘recklessly disregarded’ the possibility that ‘its conduct represented infringement.’ ”) (quoting Hamil Am., Inc. v. GFI, Inc., 193 F.3d 92, 97 (2d Cir.1999) (additional citations omitted)); Wildlife Express Corp. v. Carol Wright Sales, 18 F.3d 502, 511-12 (7th Cir.1994) (same); RCA/Ariola Int’l, Inc. v. Thomas & Grayston Co., 845 F.2d 773, 779 (8th Cir.1988) (same); see also eBay Inc. v. MercExchange, L.L.C., — U.S. -, 126 S.Ct. 1837, 1840, 164 L.Ed.2d 641 (2006) (noting with approval that its resolution of the permanent injunction standard in the patent context created harmony with copyright law).
Just recently, the Supreme Court addressed the meaning of willfulness as a statutory condition of civil liability for punitive damages. Safeco Ins. Co. of Am. v. Burr, — U.S. -, 127 S.Ct. 2201, 167 L.Ed.2d 1045 (2007). Safeco involved the Fair Credit Reporting Act (“FCRA”), which imposes civil liability for failure to comply with its requirements. Whereas an affected consumer can recover actual damages for negligent violations of the FCRA, 15 U.S.C. § 1681 o(a), he can also recover punitive damages for willful ones, 15 U.S.C. § 1681 n(a). Addressing the [1371] willfulness requirement in this context, the Court concluded that the “standard civil usage” of “willful” includes reckless behavior. Id. at 2209; accord McLaughlin v. Richland Shoe Co., 486 U.S. 128, 132-33, 108 S.Ct. 1677, 100 L.Ed.2d 115 (1988) (concluding that willful violations of the Fair Labor Standards Act include reckless violations); Trans World Airlines, Inc. v. Thurston, 469 U.S. 111, 128, 105 S.Ct. 613, 83 L.Ed.2d 523 (1985). Significantly, the Court said that this definition comports with the common law usage, “which treated actions in ‘reckless disregard’ of the law as ‘willful’ violations.” Id. at 2209 (citing W. Keeton, D. Dobbs, R. Keeton, & D. Owen, Prosser and Keeton on Law of Torts § 34, p. 212 (5th ed.1984)).
In contrast, the duty of care announced in Underwater Devices sets a lower threshold for willful infringement that is more akin to negligence. This standard fails to comport with the general understanding of willfulness in the civil context, Richland Shoe Co., 486 U.S. at 133, 108 S.Ct. 1677 (“The word ‘willful’ ... is generally understood to refer to conduct that is not merely negligent.”), and it allows for punitive damages in a manner inconsistent with Supreme Court precedent, see, e.g., Safeco, 127 S.Ct. at 2209, 2214-15, 2216 n.20; Smith v. Wade, 461 U.S. 30, 39-49, 103 S.Ct. 1625, 75 L.Ed.2d 632 (1983). Accordingly, we overrule the standard set out in Underwater Devices and hold that proof of willful infringement permitting enhanced damages requires at least a showing of objective recklessness. Because we abandon the affirmative duty of due care, we also reemphasize that there is no affirmative obligation to obtain opinion of counsel.
We fully recognize that “the term [reckless] is not self-defining.” Farmer v. Brennan, 511 U.S. 825, 836, 114 S.Ct. 1970, 128 L.Ed.2d 811 (1994). However, “[t]he civil law generally calls a person reckless who acts ... in the face of an unjustifiably high risk of harm that is either known or so obvious that it should be known.” Id. (citing Prosser and Keeton § 34, pp. 213-14; Restatement (Second) of Torts § 500 (1965)). Accordingly, to establish willful infringement, a patentee must show by clear and convincing evidence that the infringer acted despite an objectively high likelihood that its actions constituted infringement of a valid patent. See Safeco, 127 S.Ct. at 2215 (“It is [a] high risk of harm, objectively assessed, that is the essence of recklessness at common law.”). The state of mind of the accused infringer is not relevant to this objective inquiry. If this threshold objective standard is satisfied, the patentee must also demonstrate that this objectively-defined risk (determined by the record developed in the infringement proceeding) was either known or so obvious that it should have been known to the accused infringer. We leave it to future cases to further develop the application of this standard.5
Finally, we reject the argument that revisiting our willfulness doctrine is either improper or imprudent, as Convolve contends. The ultimate dispute in this case is the proper scope of discovery. While it is true that the issue of willful infringement, or even infringement for that matter, has not been decided by the trial court, it is indisputable that the proper legal standard for willful infringement informs the relevance of evidence relating to that issue [1372] and, more importantly here, the proper scope of discovery. See United States Nat’l Bank of Or. v. Indep. Ins. Agents of Am., Inc., 508 U.S. 439, 447, 113 S.Ct. 2173, 124 L.Ed.2d 402 (1993) (“[A] court may consider an issue ‘antecedent to ... and ultimately dispositive of the dispute before it, even an issue the parties fail to identify and brief.” (quoting Arcadia v. Ohio Power Co., 498 U.S. 73, 77, 111 S.Ct. 415, 112 L.Ed.2d 374 (1990))); see also Fed. R. Civ. Pro. R. 26(b) (limiting discovery to relevant, not necessarily admissible, information); accord Singleton v. Wulff, 428 U.S. 106, 121, 96 S.Ct. 2868, 49 L.Ed.2d 826 (1976) (“The matter of what questions may be taken up and resolved for the first time on appeal is one left primarily to the discretion of the courts of appeals, to be exercised on the facts of individual cases.”); Forshey v. Principi, 284 F.3d 1335, 1355-59 (Fed.Cir.2002) (en banc). Accordingly, addressing willfulness is neither hypothetical nor advisory.
II. Attorney-Client Privilege
We turn now to the appropriate scope of waiver of the attorney-client privilege resulting from an advice of counsel defense asserted in response to a charge of willful infringement. Recognizing that it is “the oldest of the privileges for confidential communications known to the common law,” we are guided by its purpose “to encourage full and frank communication between attorneys and their clients and thereby promote broader public interests in the observance of law and administration of justice.” Upjohn Co. v. United States, 449 U.S. 383, 389, 101 S.Ct. 677, 66 L.Ed.2d 584 (1981). The privilege also “recognizes that sound legal advice or advocacy serves public ends and that such advice or advocacy depends upon the lawyer’s being fully informed by the client.” Id.
The attorney-client privilege belongs to the client, who alone may waive it. E.g., Knorr-Bremse, 383 F.3d at 1345; Am. Standard, Inc. v. Pfizer, Inc., 828 F.2d 734, 745 (Fed.Cir.1987). “The widely applied standard for determining the scope of a waiver ... is that the waiver applies to all other communications relating to the same subject matter.” Fort James Corp. v. Solo Cup Co., 412 F.3d 1340, 1349 (Fed.Cir.2005). This broad scope is grounded in principles of fairness and serves to prevent a party from simultaneously using the privilege as both a sword and a shield; that is, it prevents the inequitable result of a party disclosing favorable communications while asserting the privilege as to less favorable ones. EchoStar, 448 F.3d at 1301; Fort James, 412 F.3d at 1349. Ultimately, however, “[t]here is no bright line test for determining what constitutes the subject matter of a waiver, rather courts weigh the circumstances of the disclosure, the nature of the legal advice sought and the prejudice to the parties of permitting or prohibiting further disclosures.” Fort James, 412 F.3d at 1349-50.
In considering the scope of waiver resulting from the advice of counsel defense, district courts have reached varying results with respect to trial counsel. Some decisions have extended waiver to trial counsel, e.g., Informatica Corp. v. Bus. Objects Data Integration, Inc., 454 F.Supp.2d 957 (N.D.Cal.2006), whereas others have declined to do so, e.g., Collaboration Props., Inc. v. Polycom, Inc., 224 F.R.D. 473, 476 (N.D.Cal.2004); Ampex Corp. v. Eastman Kodak Co., 2006 WL 1995140, 2006 U.S. Dist. LEXIS 48702 (D.Del. July 17, 2006). Still others have taken a middle ground and extended waiver to trial counsel only for communications contradicting or casting doubt on the opinions asserted. E.g., Intex Recreation [1373] Corp. v. Team Worldwide Corp., 439 F.Supp.2d 46 (D.D.C.2006); Beneficial Franchise Co., Inc. v. Bank One, N.A., 205 F.R.D. 212 (N.D.Ill.2001); Micron Separations, Inc. v. Pall Corp., 159 F.R.D. 361 (D.Mass.1995).
Recognizing the value of a common approach and in light of the new willfulness analysis set out above, we conclude that the significantly different functions of trial counsel and opinion counsel advise against extending waiver to trial counsel. Whereas opinion counsel serves to provide an objective assessment for making informed business decisions, trial counsel focuses on litigation strategy and evaluates the most successful manner of presenting a case to a judicial decision maker. And trial counsel is engaged in an adversarial process. We previously recognized this distinction with respect to our prior willfulness standard in Crystal Semiconductor Corp. v. TriTech Microelectronics International, Inc., 246 F.3d 1336, 1352 (Fed.Cir.2001), which concluded that “defenses prepared [by litigation counsel] for a trial are not equivalent to the competent legal opinion of non-infringement or invalidity which qualify as ‘due care’ before undertaking any potentially infringing activity.” Because of the fundamental difference between these types of legal advice, this situation does not present the classic “sword and shield” concerns typically mandating broad subject matter waiver. Therefore, fairness counsels against disclosing trial counsel’s communications on an entire subject matter in response to an accused in-fringer’s reliance on opinion counsel’s opinion to refute a willfulness allegation.
Moreover, the interests weighing against extending waiver to trial counsel are compelling. The Supreme Court recognized the need to protect trial counsel’s thoughts in Hickman v. Taylor, 329 U.S. 495, 510-11, 67 S.Ct. 385, 91 L.Ed. 451 (1947):
[I]t is essential that a lawyer work with a certain degree of privacy, free from unnecessary intrusion by opposing parties and their counsel. Proper preparation of a client’s case demands that he assemble information, sift what he considers to be the relevant from the irrelevant facts, prepare his legal theories and plan his strategy without undue and needless interference. That is the historical and the necessary way in which lawyers act within the framework of our system of jurisprudence to promote justice and to protect their clients’ interests.
The Court saw that allowing discovery of an attorney’s thoughts would result in “[ijnefficiency, unfairness and sharp practices,” that “[t]he effect on the legal profession would be demoralizing” and thus “the interests of the clients and the cause of justice would be poorly served.” Id. at 511, 67 S.Ct. 385. Although Hickman concerned work product protection, the attorney-client privilege maintained with trial counsel raises the same concerns in patent litigation. In most cases, the demands of our adversarial system of justice will far outweigh any benefits of extending waiver to trial counsel. See Jaffee v. Redmond, 518 U.S. 1, 9, 116 S.Ct. 1923, 135 L.Ed.2d 337 (1996) (“Exceptions from the general rule disfavoring testimonial privileges may be justified, however, by a ‘public good transcending the normally predominant principle of utilizing all rational means for ascertaining the truth.’ ”) (quoting Trammel v. United States, 445 U.S. 40, 50, 100 S.Ct. 906, 63 L.Ed.2d 186 (1980) (quoting Elkins v. United States, 364 U.S. 206, 80 S.Ct. 1437, 4 L.Ed.2d 1669 (1960) (Frank[1374] furter, J., dissenting))) (additional internal quotation marks omitted).
Further outweighing any benefit of extending waiver to trial counsel is the realization that in ordinary circumstances, willfulness will depend on an in-fringer’s prelitigation conduct. It is certainly true that patent infringement is an ongoing offense that can continue after litigation has commenced. However, when a complaint is filed, a patentee must have a good faith basis for alleging willful infringement. Fed. R. Civ. Pro. 8, 11(b). So a willfulness claim asserted in the original complaint must necessarily be grounded exclusively in the accused infringer’s pre-filing conduct. By contrast, when an accused infringer’s post-filing conduct is reckless, a patentee can move for a preliminary injunction, which generally provides an adequate remedy for combating post-filing willful infringement. See 35 U.S.C. § 283; Amazon.com, Inc. v. Barnesandnoble.com, Inc., 239 F.3d 1343, 1350 (Fed.Cir. 2001). A patentee who does not attempt to stop an accused infringer’s activities in this manner should not be allowed to accrue enhanced damages based solely on the infringer’s post-filing conduct. Similarly, if a patentee attempts to secure in-junctive relief but fails, it is likely the infringement did not rise to the level of recklessness.
We fully recognize that an accused in-fringer may avoid a preliminary injunction by showing only a substantial question as to invalidity, as opposed to the higher clear and convincing standard required to prevail on the merits. Amazon.com, 239 F.3d at 1359 (“Vulnerability is the issue at the preliminary injunction stage, while validity is the issue at trial. The showing of a substantial question as to invalidity thus requires less proof than the clear and convincing showing necessary to establish invalidity itself.”). However, this lessened showing simply accords with the requirement that recklessness must be shown to recover enhanced damages. A substantial question about invalidity or infringement is likely sufficient not only to avoid a preliminary injunction, but also a charge of willfulness based on post-filing conduct.
We also recognize that in some cases a patentee may be denied a preliminary injunction despite establishing a likelihood of success on the merits, such as when the remaining factors are considered and balanced. In that event, whether a willfulness claim based on conduct occurring solely after litigation began is sustainable will depend on the facts of each case.
Because willful infringement in the main must find its basis in prelitigation conduct, communications of trial counsel have little, if any, relevance warranting their disclosure, and this further supports generally shielding trial counsel from the waiver stemming from an advice of counsel defense to willfulness. Here, the opinions of Seagate’s opinion counsel, received after suit was commenced, appear to be of similarly marginal value. Although the reasoning contained in those opinions ultimately may preclude Seagate’s conduct from being considered reckless if infringement is found, reliance on the opinions after litigation was commenced will likely be of little significance.
In sum, we hold, as a general proposition, that asserting the advice of counsel defense and disclosing opinions of opinion counsel do not constitute waiver of the attorney-client privilege for communications with trial counsel. We do not purport to set out an absolute rule. Instead, trial courts remain free to exercise their [1375] discretion in unique circumstances to extend waiver to trial counsel, such as if a party or counsel engages in chicanery. We believe this view comports with Supreme Court precedent, which has made clear that rules concerning privileges are subject to review and revision, when necessary. See Jaffee, 518 U.S. at 9, 116 S.Ct. 1923 (noting that federal courts are “to ‘continue the evolutionary development of testimonial privileges.’ ” (quoting Trammel, 445 U.S. at 47, 100 S.Ct. 906)).
III. Work Product Protection
An advice of counsel defense asserted to refute a charge of willful infringement may also implicate waiver of work product protection. Again, we are here confronted with whether this waiver extends to trial counsel’s work product. We hold that it does not, absent exceptional circumstances.
The work product doctrine is “designed to balance the needs of the adversary system: promotion of an attorney’s preparation in representing a client versus society’s general interest in revealing all true and material facts to the resolution of a dispute.” In re Martin Marietta Corp., 856 F.2d 619, 624 (4th Cir.1988). Unlike the attorney-client privilege, which provides absolute protection from disclosure, work product protection is qualified and may be overcome by need and undue hardship. Fed. R. Civ. Pro. 26(b)(3). However, the level of need and hardship required for discovery depends on whether the work product is factual, or the result of mental processes such as plans, strategies, tactics, and impressions, whether memorialized in writing or not. Whereas factual work product can be discovered solely upon a showing of substantial need and undue hardship, mental process work product is afforded even greater, nearly absolute, protection. See id.; Upjohn Co. v. United States, 449 U.S. 383, 400, 101 S.Ct. 677, 66 L.Ed.2d 584 (1981); Holmgren v. State Farm Mut. Auto. Ins., 976 F.2d 573, 577 (9th Cir.1992) (holding that work product “may be discovered and admitted when mental impressions are at issue in a case and the need for the material is compelling”); see also Director, Office of Thrift Supervision v. Vinson & Elkins, LLP, 124 F.3d 1304, 1307 (D.C.Cir.1997) (“virtually undiseoverable”). But see Nat’l Union Fire Ins. Co. v. Murray Sheet Metal Co., 967 F.2d 980, 984 (4th Cir.1992) (“ ‘absolutely’ immune from discovery”).
Like the attorney-client privilege, however, work product protection may be waived. United States v. Nobles, 422 U.S. 225, 239, 95 S.Ct. 2160, 45 L.Ed.2d 141 (1975). Here, the same rationale generally limiting waiver of the attorney-client privilege with trial counsel applies with even greater force to so limiting work product waiver because of the nature of the work product doctrine. Protecting lawyers from broad subject matter of work product disclosure “strengthens the adversary process, and ... may ultimately and ideally further the search for the truth.” Martin Marietta, 856 F.2d at 626; accord EchoStar, 448 F.3d at 1301 (“[W]ork-product immunity ... promotes a fair and efficient adversarial system.... ”); Coastal States Gas Corp. v. Dep’t of Energy, 617 F.2d 854, 864 (D.C.Cir.1980) (“The purpose of the privilege, however, is not to protect any interest of the attorney ... but to protect the adversary trial process itself. It is believed that the integrity of our system would suffer if adversaries were entitled to probe each other’s thoughts and plans concerning the case.”). In addition, trial counsel’s mental processes, which fall [1376] within Convolve’s discovery requests, enjoy the utmost protection from disclosure; a scope of waiver commensurate with the nature of such heightened protection is appropriate. See Martin Marietta, 856 F.2d at 625-26.
The Supreme Court has approved of narrowly restricting the scope of work product waiver. In United States v. Nobles, a criminal case, an accused armed robber presented the testimony of an investigator in an attempt to discredit the two eyewitnesses. When they testified for the prosecution, the defense attorney relied on the investigator’s report in cross-examining the eyewitnesses. 422 U.S. at 227, 95 S.Ct. 2160. After the prosecution rested, the defense attempted to call the investigator to testify. The trial court, however, ruled that if the investigator testified, his affirmative testimony would mandate disclosure of the portions of his report relating to his testimony. Id. at 229, 95 S.Ct. 2160. The Supreme Court agreed that the investigator’s affirmative testimony waived work product protection, but it approvingly noted the “quite limited” scope of waiver imposed by the trial court and its refusal to allow a general “fishing expedition” into the defense files or even the investigator’s report. Id. at 239-40, 95 S.Ct. 2160. Similarly, Convolve has been granted access to the materials relating to Seagate’s opinion counsel’s opinion, and he was made available for deposition. The extent of this waiver accords with the principles and spirit of Nobles.
Accordingly, we hold that, as a general proposition, relying on opinion counsel’s work product does not waive work product immunity with respect to trial counsel. Again, we leave open the possibility that situations may arise in which waiver may be extended to trial counsel, such as if a party or his counsel engages in chicanery. And, of course, the general principles of work product protection remain in force, so that a party may obtain discovery of work product absent waiver upon a sufficient showing of need and hardship, bearing in mind that a higher burden must be met to obtain that pertaining to mental processes. See Fed. R. Civ. Pro. 26(b)(3).
Finally, the work product doctrine was partially codified in Rule 26(b)(3) of the Federal Rules of Civil Procedure, which applies work product protection to “documents and tangible things.” Courts continue to apply Hickman v. Taylor, 329 U.S. 495, 67 S.Ct. 385, 91 L.Ed. 451, to “nontangible” work product. See, e.g., In re Cendant Corp. Sec. Litig., 343 F.3d 658, 662-63 (3d Cir.2003); United States v. One Tract of Real Property, 95 F.3d 422, 428 n. 10 (6th Cir.1996). This is relevant here because Convolve sought to depose Sea-gate’s trial counsel. We agree that work product protection remains available to “nontangible” work product under Hickman. Otherwise, attorneys’ files would be protected from discovery, but attorneys themselves would have no work product objection to depositions.
Conclusion
Accordingly, Seagate’s petition for a writ of mandamus is granted, and the district court will reconsider its discovery orders in light of this opinion.