Federal Rule of Civil Procedure 26 is a comprehensive rule — it covers initial disclosure obligations, the scope of permissible discovery, expert witness disclosure, and the duty to supplement. For e-discovery practitioners, three specific subsections drive most of the day-to-day TAR practice decisions: Rule 26(b)(1)'s proportionality framework, Rule 26(f)'s meet-and-confer requirement, and Rule 26(g)'s certification obligation. Understanding how courts have read these provisions in TAR contexts is not a theoretical exercise — it determines what your protocol documentation needs to say and how your review team's decisions get scrutinized on a motion to compel.
Rule 26(b)(1): Proportionality and TAR
The 2015 amendments to Rule 26(b)(1) placed proportionality explicitly in the scope of discovery provision. The rule now requires that discovery be "proportional to the needs of the case, considering the importance of the issues at stake in the action, the amount in controversy, the parties' relative access to relevant information, the parties' resources, the importance of the discovery in resolving the issues, and whether the burden or expense of the proposed discovery outweighs its likely benefit."
In the TAR context, this cuts both ways. Courts have used proportionality to limit discovery demands that would require processing unreasonably large ESI populations. But proportionality has also been invoked by requesting parties arguing that the producing party's keyword-only review methodology was insufficiently rigorous for the importance of the discovery — creating an affirmative argument for more, not less, sophisticated review methods.
In Hyles v. New York City (S.D.N.Y. 2016), Judge Peck addressed a requesting party's attempt to compel the producing party to use TAR rather than keyword review. Peck declined to order the switch — holding that the producing party has the right to select its own review methodology — but was explicit that the producing party would remain responsible for producing all responsive documents regardless of methodology choice. The opinion reinforces the principle that methodology choice is a producing-party decision, but the obligation to produce responsive, non-privileged documents is methodology-neutral.
The proportionality framework creates a practical planning point for TAR protocol design: your review process documentation should reflect not only how you chose your methodology but why, with reference to the factors enumerated in 26(b)(1). A proportionality narrative — here is why this TAR approach is appropriate for the scope and stakes of this matter — is increasingly standard in well-drafted ESI protocols.
Rule 26(f): The Meet-and-Confer Obligation
Rule 26(f) requires that the parties confer before a scheduling conference on, among other things, "any issues about disclosure, discovery, or preservation of electronically stored information, including the form or forms in which it should be produced." The rule does not mention predictive coding, TAR protocols, or machine learning classifiers — those specifics are creatures of case law and party practice, not the rule text.
The practical question is how specifically the parties need to disclose their TAR methodology during the 26(f) conference. Courts have varied in their expectations. At one end, the parties in Da Silva Moore v. Publicis Groupe (S.D.N.Y. 2012) negotiated a detailed TAR protocol with the court's involvement, including specifications for the seed set composition, review tiers, and validation thresholds. At the other end, many courts have been satisfied with ESI protocols that specify the general methodology (technology assisted review / predictive coding) without requiring detailed technical parameters at the meet-and-confer stage.
The Sedona Conference Cooperation Proclamation and the more detailed guidance in the Sedona Principles (3rd edition, Principle 6) encourage parties to cooperate on discovery process decisions, including TAR methodology, as a way of reducing motion practice about discovery disputes. Cooperation in this context does not mean agreement — it means disclosure and good-faith engagement with the opposing party's questions about your methodology.
Consider a practical scenario: discovery counsel for a defendant in a commercial dispute discloses at the 26(f) conference that the company intends to use predictive coding for its document review. Plaintiff's counsel immediately asks: what platform, what seed set size, what recall threshold? The producing party's counsel is not obligated to answer in technical detail at that stage — but the manner and completeness of the response sets the tone for the rest of the discovery process. Counsel who can speak competently to the methodology tend to face fewer contested motions than counsel who respond with vague assurances about "AI review."
Rule 26(g): Certification and Its Implications for TAR
Rule 26(g) requires that every discovery disclosure, discovery request, or discovery response be signed by an attorney who, after reasonable inquiry, certifies that the disclosure or response is complete and correct and that any discovery response is not unduly burdensome or expensive in light of the factors listed in Rule 26(b)(1).
The "reasonable inquiry" standard under 26(g) is where TAR methodology transparency becomes a professional obligation rather than a strategic choice. The certifying attorney does not need to have designed the TAR protocol — but they need to understand it well enough to vouch for the completeness of the production. Courts have expected certifying attorneys to be able to describe, at a basic level, the review methodology used, what validation was performed, and why the certification of completeness is supportable.
This has two implications. First, the TAR protocol documentation must be accessible to and understandable by the certifying attorney — not just the e-discovery vendor or the litigation support director who designed it. Protocol documentation written in technical jargon that the signing attorney cannot interpret creates a 26(g) certification problem. Second, the elusion testing results (the sampling validation that confirms the non-reviewed population does not contain a statistically significant number of responsive documents) must be documented in a form the certifying attorney can rely on.
We're not saying the certifying attorney needs to become a TAR expert — that is not a realistic expectation, nor is it what Rule 26(g) requires. The standard is reasonable inquiry, not technical expertise. But the attorney's certification is only as solid as the documentation underlying it, and protocol documentation designed for attorney comprehension — not just vendor defensibility — is a meaningful difference in practice.
ESI Protocol Structure for TAR Matters
A well-structured ESI protocol in a matter using TAR should address the following elements, most of which map directly to the Rule 26 provisions discussed above:
- Scope definition: Custodians, date ranges, data sources (email, file share, collaboration tools, mobile). This defines the corpus on which TAR will be applied and supports the proportionality narrative under 26(b)(1).
- Methodology identification: SPL/TAR 1.0 or CAL/TAR 2.0, naming the platform or classifier family. Satisfies the 26(f) disclosure expectation.
- Training process: Seed set or initial training document composition, including how the initial training corpus was assembled and by whom.
- Validation methodology: Elusion testing protocol, target recall threshold, sampling methodology (random, stratified), and what happens if validation fails to meet the threshold on the first pass.
- Privilege review integration: How privilege-flagged documents are separated from the responsive/non-responsive determination workflow, and what attorney review tier handles them.
- Production format: TIFF with load file, native, or hybrid; metadata fields included (consistent with the standard fields expected in DAT/Concordance or .OPT image load file conventions); Bates numbering scheme.
- Dispute resolution: What process the parties will follow if the requesting party disputes the producing party's recall threshold or methodology decisions.
When Opposing Counsel Challenges Your TAR Process
TAR-related motions to compel typically do not argue that the methodology was impermissible in principle — the judicial acceptance of TAR since Da Silva Moore has made that argument largely unavailable. Instead, opposing counsel challenges tend to focus on: (1) whether the producing party followed the protocol they agreed to; (2) whether the validation metrics actually support the certification of completeness; and (3) whether the privilege log generated by the automated review is legally sufficient.
The most effective defense against all three challenges is process documentation that was built from the protocol stage, not reconstructed after a motion is filed. Counsel who can produce contemporaneous logs of seed set review decisions, training round timestamps, and elusion test results are in a fundamentally different position than counsel who can only offer a vendor's post-hoc summary of what the platform did.
Protocol design for TAR matters requires legal judgment that cannot be reduced to a software configuration. If you'd like to discuss how defensible review workflows are structured for the specific requirements of your practice, the walkthrough request form is a starting point.