社会ガイド
Predictive Coding and Court Approval: Da Silva Moore and After
The 2012 Da Silva Moore decision approved a negotiated predictive-coding protocol for one large electronic discovery matter.
このページでは3 分で読めます
概要
It did not endorse a vendor, require predictive coding in every case, or guarantee accuracy; it emphasized transparency, quality control, and context-specific review.
ディープダイブ
Predictive coding, also called technology-assisted review or computer-assisted review, uses human judgments about sample documents to train or guide software in prioritizing potentially relevant electronically stored information (ESI). In Da Silva Moore v. Publicis Groupe, the U.S. District Court for the Southern District of New York considered a parties’ negotiated ESI protocol in a large discovery matter. The February 2012 order allowed computer-assisted review under that protocol, with iterative review and transparency provisions. The decision was an early judicial approval in its factual setting, not a rule that predictive coding must be used or that a particular vendor or tool is endorsed. The court emphasized circumstances including the parties’ agreement, large document volume, available alternatives, cost and proportionality, and the proposed process’s transparency. A later district-judge order affirmed the magistrate judge’s discovery ruling under the applicable deferential review standard. Neither order declared the software infallible; the record recognized that no review method guarantees perfection. For modern discovery, counsel should agree on scope, responsiveness criteria, training and validation samples, privilege handling, production format, and quality control where appropriate. Monitor recall and precision as useful measures, but explain sampling limits and the consequences of errors. The case does not decide every future protocol or replace the Federal Rules, local rules, or case-specific judicial discretion. Treat it as an example of a court evaluating a transparent, negotiated protocol, not as blanket approval of any automated review process. In practice, the parties should clarify what counts as responsive before training, and specify how privileged material will be screened. Quality control can include a second-level review of selected nonresponsive documents and additional review when sampling reveals uncertainty. A negotiated protocol should explain who receives information about the process and how disputes are brought to the court.
戦略的影響
リスクと安全性
AI による壊滅的な被害も日常的な被害も、誰がリスクを理解し、誰が行動できるかにかかっています。
より明確な判決
国民と専門家のリテラシーは、強力な安全政策が政治的に可能かどうかを左右します。
誇大広告を打ち破る
明確な説明は、誇大広告、研究室の PR、曖昧な倫理劇場に囚われることを減らします。
The Future of Predictive Coding and Court Approval: Da Silva Moore and After
Courts continue to encounter different data volumes, tools, and party agreements. Predictive coding remains one option within technology-assisted review, with defensibility depending on the protocol and case record. Teams should document methods and adapt validation to the matter rather than relying on a single historical decision as a guarantee. Later cases may take different approaches because the data, parties, rules, and protocols differ. Counsel should use current rules and local precedent, consult on proportionality, and avoid presenting one district court order as a nationwide mandate. The enduring lesson is to make the method and quality controls explainable in the specific matter.
現実世界の実装
Counsel proposes an ESI protocol and documents training and quality-control steps.
A discovery team discusses sampling and transparent reporting with opposing counsel.
A lawyer avoids citing Da Silva Moore as a universal mandate for predictive coding.
A court evaluates proportionality and the record in the case before it.
リスクとガードレール
能力が複雑になる一方で、実存的なリスクを SF として扱います。
高度な自律性の下での調整による表面製品の安全性を混乱させる。
英語以外や専門家ではない聴衆には、低品質の情報源しか提供されません。
実装ロードマップ
製品の危害、誤使用、制御不能/調整不良のリスクを分離します。
どのような証拠がタイムラインと重大度についてのあなたの見方を変えるかを尋ねてください。
マーケティング上の主張よりも、一次情報源と具体的な評価を優先します。
意識だけでなく、キャリア、政策、資金、スキルなど、行動経路を 1 つ特定します。
探検を続けましょう
Free newsletter
Get the daily AI briefing
Three verified AI stories every weekday morning, written in plain English. Free forever, no ads.
One email each weekday. Unsubscribe in one click. We never sell or share your address.
Test yourself
Take the Predictive Coding and Court Approval: Da Silva Moore and After quiz
Instant feedback on every answer, and a shareable certificate with a verifiable ID once you pass a course.
Support free AI education. AI Understanding is a 501(c)(3) nonprofit — no ads, no paywall, ever. Make a donation
よくある質問
What is Predictive Coding and Court Approval: Da Silva Moore and After?
The 2012 Da Silva Moore decision approved a negotiated predictive-coding protocol for one large electronic discovery matter. It did not endorse a vendor, require predictive coding in every case, or guarantee accuracy; it emphasized transparency, quality control, and context-specific review.
What did the Da Silva Moore court approve?
The opinion approved the case-specific protocol and explicitly did not mandate use in every case.
Which factor supported approval in the case?
The court considered agreement, volume, alternatives, proportionality, and transparency.
What does the case imply about future matters?
The opinion cautions that this protocol is not automatically right for future cases.
Why are transparency and quality control important in a review protocol?
The court considered transparent process and quality review, while recognizing no tool is perfect.
In e-discovery, which workflow is called predictive coding in the guide?
Predictive coding supports review of electronic documents; legal relevance criteria remain contextual.
学び続ける
関連ガイド
このトピックのために選ばれたその他のガイド