Privileges, Clawbacks, and Inadvertent Disclosures

Technology-Assisted Review Is Judicially Approved

Technology'assisted review software goes by many names, including “computer assisted review,” “predictive coding” and “machine learning.” Whatever its moniker, the software, using coding input by humans (lawyers), reviews or scans the universe of potentially relevant documents and identifies those documents most likely to contain relevant information. The initial step calls for the company lawyer to review and code a seed set of relevant documents. The TAR software identifies properties of the seed documents and uses these properties to review the entire set of potentially relevant documents and identify only the most relevant. The software may categorize documents as relevant in a yes/no format or by a relevance range (e.g., a score of 15 on a 100 scale is less relevant, while a score of 85 on a 100 scale is highly relevant).

TAR software has advantages over the traditional linear document review. Statistics show that one advantage is that TAR review identifies relevant documents more accurately than human review. Another advantage is that utilization of TAR software greatly reduces attorney time and fees associated with a human review of the entire document universe. Because of its greater accuracy at a significantly lower cost, many corporate counsel have clamored for greater use of this emerging technology.

But these same counsel were fearful that TAR-related efforts would prove futile unless understood and approved by trial judges. And in the February 2012 decision in Da Silva Moore v. Publicis Groupe, 2012 WL 607412 (S.D.N.Y. Feb. 24, 2012), Magistrate Judge Andrew J. Peck became the first to judicially approve, in a written opinion, the use of TAR in the e-discovery context. Judge Peck, a long-time proponent of TAR software, noted that statistics show that computerized searches are at least as accurate as manual review. The court also noted that TAR software yielded much more effective results than traditional keyword searches, which he equated with the children's game “Go Fish!” And noting that TAR reviews require only 1.9% of human review of documents, he said the savings associated with privilege review were too drastic to ignore.

TAR software is not useful in all instances, providing its greatest benefit in large document cases. But with Judge Peck's influence in the e-discovery world, many believe that his Da Silva Moore decision will serve as a catalyst for greater use of TAR software in appropriate situations. The question whether to utilize a TAR application should hinge not simply on the volume of documents, but also on how the software can solve corporate counsel's Catch-22 dilemma of controlling privilege review costs while ensuring that privileged documents are not inadvertently disclosed.

Protection of Evidentiary Privileges

Corporate counsel's concerns about releasing privileged information go hand-in-hand with the rise in the volume of electronically stored information. The greater the volume of documents produced, the greater the chance that privileged information is inadvertently released. Courts quickly realized that e-discovery costs were increasing at a prohibitive rate and that attorneys' fees associated with privilege review was a significant portion of those costs. To remedy the growing problem, courts sought to reduce privilege review through the use of lenient inadvertent disclosure rules and nonwaiver or clawback agreements, primary through Federal Rule of Evidence 502.

A clawback agreement is effectively an agreement between the parties that, if certain defined protocols are followed, the disclosure of privileged information within a document production will not automatically constitute a waiver of the privilege. Federal Rule of Evidence 502 approves of clawback agreements as a way to avoid privilege waiver, and seeks to extend the nonwaiver doctrine to subsequent litigation in both federal and state courts. Rule 502(e) permits clawback agreements to govern waiver between the parties, but also to nonparties if incorporated into a Protective Order. And, with little regard for the Erie doctrine or state automony, Rule 502(f) extends the nonwaiver protection to subsequent proceedings whether in federal or state court.

Rule 502, however, has not turned out to be the cost-saving answer that many thought. The rule only applies to nonwaiver of the attorney-client privilege and the work product doctrine. It completely ignores that companies maintain electronically stored information that may be protected by other evidentiary privileges. For example, a healthcare provider likely maintains information protected by a state-law peer-review privilege; and other businesses maintain information protected by the state-law accountant-client privilege. And whether a Rule 502 sanctioned clawback agreement is enforceable in subsequent state-court proceedings remains unclear, but a likely scenario is that state courts will find privilege waiver regardless of any federal clawback agreement. In short, Rule 502 has limited application and uncertain enforceability that should give corporate counsel pause before relying on clawback agreements as a cost-saving tool.

To be sure, courts' embrace of clawback agreements as an e-discovery cost-reduction method has some basis of support. Some statistics show that these agreements cut privilege-review costs by 50%. Yet, others argue that these costs are not reduced but simply shifted to the opposing party; the producing party's review time and associated costs decrease but the receiving party's review time and costs correspondingly increase. This phenomenon is especially true in commercial litigation where both parties maintain large volume of potentially relevant documents.

Beyond the questionable cost savings, clawback agreements have not proven to be attractive options in practice. Clawback agreements do not obviate lawyer's ethical duty, exemplified in Model Rule 1.6, to competently safeguard against the unauthorized disclosure of her client's confidential information. Even with a clawback agreement in place, many courts have found privilege waiver where the producing party did not act with at least some diligence to identify and withhold privileged documents. And even when enforced, there is no guarantee that a nonwaiver finding in one proceeding results in similar nonwaiver finding in a subsequent proceeding. Consequently, clawback agreements do not necessarily safeguard a client's confidential information, and lawyers relying on them to do so run the risk of breaching state ethical obligations.

A more practical concern with clawback agreements is that a less than thorough privilege review will inevitably result in the production of privileged information. And even when a clawback agreement works perfectly, it is difficult to assess the damage done by permitting opposing counsel to review privileged information. Using whatever clich' you prefer, once privileged information is disclosed, you cannot “unring the bell,” “put the toothpaste back in the tube,” or “put the genie back in the bottle.” In short, you cannot erase privileged information from your adversary's mind, and even when the privileged documents are returned, you can rarely assess how your adversary's review of the information will alter his litigation strategy. Will the information cause your adversary to depose new witnesses, research new legal theories not previously considered, or send discovery into a new subject area? So, while a clawback agreement may, in theory, reduce your client's e-discovery costs, it is difficult to quantify the added costs ' including an adverse verdict ' that may result from the disclosure of privileged information.

TAR to the Rescue?

So how can technology-assisted review software better solve corporate counsel's legitimate concerns over the issue of reducing costs associated with attorneys' privilege review while taking greater measures to prevent the disclosure of privileged information? There are two ways in which this new technology can alleviate corporate counsel's concerns.

First, TAR software should provide cost-savings that equal or exceed the cost-savings most associate with clawback agreements, but without the risks arising from inadvertent disclosure. Noting that the most significant portion of e-discovery costs come from attorneys' review for privileged information, clawback agreements seek to achieve savings in this area by encouraging less attorney review time. But TAR software will more accurately identify the scope of relevant documents for production and reduce the number of irrelevant documents. The result is that the scope of documents ultimately submitted for attorneys' privilege review is smaller which necessary means that attorney review time is correspondingly shorter. Thus, by using TAR to identify relevant documents, corporate counsel can afford to have a thorough privilege review, completed in less time, and achieve the same cost-savings as she would by using a clawback agreement and a less thorough review. And the cost-savings will be achieved without the ethical and strategy-related problems that arise from clawback agreements and inadvertent disclosures.

Second, while TAR software was designed to take coding and data entered by the client's lawyer and search for electronically stored information responsive to an adversary's discovery request, the same software can be used to identify and separate the privileged information within the company's electronic storage areas. For example, company counsel may identify keyword searches that would likely return privileged information. For attorney-client privileged information, the keywords may be the lawyers' names or words associated with the claim at issue; for peer-review privileged information the keywords may be associated with quality and assessment of care. Whatever the identifiers, the returned information will be used as a seed set for coding, and the TAR software will then scan the remaining universe of documents to locate those most likely to contain privileged information. Corporate counsel can save significant attorney-review costs by simply having the computer review the documents instead. And studies show that the software will actually be more accurate in identifying privileged information, thereby further reducing the possibility of inadvertent disclosure.

Summaries and Practical Tips

With reduction of e-discovery costs and privilege protection paramount considerations for corporate counsel, the following practical tips and summaries should provide focus and alleviate concerns.

• The Da Silva Moore decision approving the use of technology-assisted review software will likely increase corporate counsel's desire to utilize the software and achieve greater accuracy at lower costs. Corporate counsel should determine whether this technology is appropriate for her cases.
• Corporate counsel should be aware that inadvertent disclosure rules are not applied uniformly and do not always provide satisfactory safeguards against privilege waiver.
• Corporate counsel should be leery of using clawback agreements. Heralded by some as a protective umbrella, these agreements may create a conflict with counsel's ethical obligations and provide false assurances that nonwaiver rulings will apply in subsequent litigation.
• Corporate counsel must diligently protect against the disclosure of privileged information, and the legitimate goals of saving on e-discovery expenses should not impair counsel's diligence.

Conclusion

In considering whether technology-assisted software is the right method to identify and collect electronically stored information, corporate counsel should consider how this technology may make privilege review more accurate and allow for a more thorough attorney review at a more efficient cost.

Several issues and concerns populate corporate counsels' minds when confronted with e-discovery demands, but two rise to the top: 1) collection and production cost; and 2) inadvertently producing information protected by evidentiary privileges. And these two concerns overlap, producing a Catch-22 dilemma for in-house lawyers. On the one hand, a complete pre-production privilege review of documents constitutes the highest-cost item in the e-discovery process. On the other hand, producing documents with limited to no privilege review risks inadvertent disclosures and privilege waiver, which may result is disastrous ethical and strategy-related consequences.

Many thought that courts' approval of lenient inadvertent disclosure rules and nonwaiver, or clawback, agreements would solve this double-edged problem. But neither avenue proved acceptable in practice. Some courts still found privilege waiver when documents were mistakenly disclosed without acceptable safeguards; and other courts ruled that clawback agreements did not obviate a pre-production privilege review and were not enforceable against third parties in subsequent litigation. And in these situations, parties were neither saving production-associated costs nor protecting their privileged information from discovery.

A byproduct of technology-assisted review (TAR) software may provide the cost-savings and privilege protections that have eluded corporate counsel during the e-discovery age. This software, designed to identify relevant discovery documents without extensive attorney review, will more readily identify privileged documents and narrow the universe of relevant documents, thereby reducing costs associated with attorneys' privilege review in a way that inadvertent disclosure rules or clawback agreements never could.

Technology-Assisted Review Is Judicially Approved

Technology'assisted review software goes by many names, including “computer assisted review,” “predictive coding” and “machine learning.” Whatever its moniker, the software, using coding input by humans (lawyers), reviews or scans the universe of potentially relevant documents and identifies those documents most likely to contain relevant information. The initial step calls for the company lawyer to review and code a seed set of relevant documents. The TAR software identifies properties of the seed documents and uses these properties to review the entire set of potentially relevant documents and identify only the most relevant. The software may categorize documents as relevant in a yes/no format or by a relevance range (e.g., a score of 15 on a 100 scale is less relevant, while a score of 85 on a 100 scale is highly relevant).

TAR software has advantages over the traditional linear document review. Statistics show that one advantage is that TAR review identifies relevant documents more accurately than human review. Another advantage is that utilization of TAR software greatly reduces attorney time and fees associated with a human review of the entire document universe. Because of its greater accuracy at a significantly lower cost, many corporate counsel have clamored for greater use of this emerging technology.

But these same counsel were fearful that TAR-related efforts would prove futile unless understood and approved by trial judges. And in the February 2012 decision in Da Silva Moore v. Publicis Groupe, 2012 WL 607412 (S.D.N.Y. Feb. 24, 2012), Magistrate Judge Andrew J. Peck became the first to judicially approve, in a written opinion, the use of TAR in the e-discovery context. Judge Peck, a long-time proponent of TAR software, noted that statistics show that computerized searches are at least as accurate as manual review. The court also noted that TAR software yielded much more effective results than traditional keyword searches, which he equated with the children's game “Go Fish!” And noting that TAR reviews require only 1.9% of human review of documents, he said the savings associated with privilege review were too drastic to ignore.

TAR software is not useful in all instances, providing its greatest benefit in large document cases. But with Judge Peck's influence in the e-discovery world, many believe that his Da Silva Moore decision will serve as a catalyst for greater use of TAR software in appropriate situations. The question whether to utilize a TAR application should hinge not simply on the volume of documents, but also on how the software can solve corporate counsel's Catch-22 dilemma of controlling privilege review costs while ensuring that privileged documents are not inadvertently disclosed.

Protection of Evidentiary Privileges

Corporate counsel's concerns about releasing privileged information go hand-in-hand with the rise in the volume of electronically stored information. The greater the volume of documents produced, the greater the chance that privileged information is inadvertently released. Courts quickly realized that e-discovery costs were increasing at a prohibitive rate and that attorneys' fees associated with privilege review was a significant portion of those costs. To remedy the growing problem, courts sought to reduce privilege review through the use of lenient inadvertent disclosure rules and nonwaiver or clawback agreements, primary through Federal Rule of Evidence 502.

A clawback agreement is effectively an agreement between the parties that, if certain defined protocols are followed, the disclosure of privileged information within a document production will not automatically constitute a waiver of the privilege. Federal Rule of Evidence 502 approves of clawback agreements as a way to avoid privilege waiver, and seeks to extend the nonwaiver doctrine to subsequent litigation in both federal and state courts. Rule 502(e) permits clawback agreements to govern waiver between the parties, but also to nonparties if incorporated into a Protective Order. And, with little regard for the Erie doctrine or state automony, Rule 502(f) extends the nonwaiver protection to subsequent proceedings whether in federal or state court.

Rule 502, however, has not turned out to be the cost-saving answer that many thought. The rule only applies to nonwaiver of the attorney-client privilege and the work product doctrine. It completely ignores that companies maintain electronically stored information that may be protected by other evidentiary privileges. For example, a healthcare provider likely maintains information protected by a state-law peer-review privilege; and other businesses maintain information protected by the state-law accountant-client privilege. And whether a Rule 502 sanctioned clawback agreement is enforceable in subsequent state-court proceedings remains unclear, but a likely scenario is that state courts will find privilege waiver regardless of any federal clawback agreement. In short, Rule 502 has limited application and uncertain enforceability that should give corporate counsel pause before relying on clawback agreements as a cost-saving tool.

To be sure, courts' embrace of clawback agreements as an e-discovery cost-reduction method has some basis of support. Some statistics show that these agreements cut privilege-review costs by 50%. Yet, others argue that these costs are not reduced but simply shifted to the opposing party; the producing party's review time and associated costs decrease but the receiving party's review time and costs correspondingly increase. This phenomenon is especially true in commercial litigation where both parties maintain large volume of potentially relevant documents.

Beyond the questionable cost savings, clawback agreements have not proven to be attractive options in practice. Clawback agreements do not obviate lawyer's ethical duty, exemplified in Model Rule 1.6, to competently safeguard against the unauthorized disclosure of her client's confidential information. Even with a clawback agreement in place, many courts have found privilege waiver where the producing party did not act with at least some diligence to identify and withhold privileged documents. And even when enforced, there is no guarantee that a nonwaiver finding in one proceeding results in similar nonwaiver finding in a subsequent proceeding. Consequently, clawback agreements do not necessarily safeguard a client's confidential information, and lawyers relying on them to do so run the risk of breaching state ethical obligations.

A more practical concern with clawback agreements is that a less than thorough privilege review will inevitably result in the production of privileged information. And even when a clawback agreement works perfectly, it is difficult to assess the damage done by permitting opposing counsel to review privileged information. Using whatever clich' you prefer, once privileged information is disclosed, you cannot “unring the bell,” “put the toothpaste back in the tube,” or “put the genie back in the bottle.” In short, you cannot erase privileged information from your adversary's mind, and even when the privileged documents are returned, you can rarely assess how your adversary's review of the information will alter his litigation strategy. Will the information cause your adversary to depose new witnesses, research new legal theories not previously considered, or send discovery into a new subject area? So, while a clawback agreement may, in theory, reduce your client's e-discovery costs, it is difficult to quantify the added costs ' including an adverse verdict ' that may result from the disclosure of privileged information.

TAR to the Rescue?

So how can technology-assisted review software better solve corporate counsel's legitimate concerns over the issue of reducing costs associated with attorneys' privilege review while taking greater measures to prevent the disclosure of privileged information? There are two ways in which this new technology can alleviate corporate counsel's concerns.

First, TAR software should provide cost-savings that equal or exceed the cost-savings most associate with clawback agreements, but without the risks arising from inadvertent disclosure. Noting that the most significant portion of e-discovery costs come from attorneys' review for privileged information, clawback agreements seek to achieve savings in this area by encouraging less attorney review time. But TAR software will more accurately identify the scope of relevant documents for production and reduce the number of irrelevant documents. The result is that the scope of documents ultimately submitted for attorneys' privilege review is smaller which necessary means that attorney review time is correspondingly shorter. Thus, by using TAR to identify relevant documents, corporate counsel can afford to have a thorough privilege review, completed in less time, and achieve the same cost-savings as she would by using a clawback agreement and a less thorough review. And the cost-savings will be achieved without the ethical and strategy-related problems that arise from clawback agreements and inadvertent disclosures.

Second, while TAR software was designed to take coding and data entered by the client's lawyer and search for electronically stored information responsive to an adversary's discovery request, the same software can be used to identify and separate the privileged information within the company's electronic storage areas. For example, company counsel may identify keyword searches that would likely return privileged information. For attorney-client privileged information, the keywords may be the lawyers' names or words associated with the claim at issue; for peer-review privileged information the keywords may be associated with quality and assessment of care. Whatever the identifiers, the returned information will be used as a seed set for coding, and the TAR software will then scan the remaining universe of documents to locate those most likely to contain privileged information. Corporate counsel can save significant attorney-review costs by simply having the computer review the documents instead. And studies show that the software will actually be more accurate in identifying privileged information, thereby further reducing the possibility of inadvertent disclosure.

Summaries and Practical Tips

With reduction of e-discovery costs and privilege protection paramount considerations for corporate counsel, the following practical tips and summaries should provide focus and alleviate concerns.

• The Da Silva Moore decision approving the use of technology-assisted review software will likely increase corporate counsel's desire to utilize the software and achieve greater accuracy at lower costs. Corporate counsel should determine whether this technology is appropriate for her cases.
• Corporate counsel should be aware that inadvertent disclosure rules are not applied uniformly and do not always provide satisfactory safeguards against privilege waiver.
• Corporate counsel should be leery of using clawback agreements. Heralded by some as a protective umbrella, these agreements may create a conflict with counsel's ethical obligations and provide false assurances that nonwaiver rulings will apply in subsequent litigation.
• Corporate counsel must diligently protect against the disclosure of privileged information, and the legitimate goals of saving on e-discovery expenses should not impair counsel's diligence.

Conclusion

In considering whether technology-assisted software is the right method to identify and collect electronically stored information, corporate counsel should consider how this technology may make privilege review more accurate and allow for a more thorough attorney review at a more efficient cost.