Tax Court lets taxpayers use predictive coding over IRS objections

In September 2014 the US Tax Court allowed predictive coding over the IRS's objection. It credited an estimate of 200,000 to 400,000 documents to review at $80,000 to $85,000.

  • Not rated an organization

    The review had not run when the court ruled, and the costs are the taxpayers' expert's estimates.

Depth runs from 1, a convenience or time saved, to 5, a life or a livelihood preserved.

Status
Intended benefit.
Evidence

foundstudiedreportedclaimed

An independent court document supports the benefit.

Compared with

Review under the Commissioner's approach, a privilege search of all restored data. Assumed without a measurement.

Without the system
3.5 million to 7 million documents at a cost of $500,000 to $550,000
With it
200,000 to 400,000 documents at a cost of $80,000 to $85,000
The system’s part
Not isolated. The court allowed predictive coding before the review ran. The cost comparison is the taxpayers' expert's estimate, which the court found credible and uncontradicted. The opinion describes software trained on a small coded sample of documents.
Limits
Not yet checked.
Costs and harms
Not yet checked.
Later
Taken to court, 13 July 2016. The review ran, and the Commissioner moved to compel more documents. On 13 July 2016 the court denied the motion and held that the response was complete.

Download data Cite Suggest a correction

Benefit type
Reduced Task Cost
System
Predictive coding
Made by
Not named in the opinion
Used by
Dynamo Holdings Limited Partnership and Beekman Vista, Inc. (petitioners)
Where
United States
Stage
Proposed
Date
, when the source was published
Kind
Benefit created
Broad benefit
Processing Information, Getting Information
Collections
Scrutiny and findings
Last revised
31 October 2026

Sources

  1. Court document, by authors with no tie to the system.

    United States Tax Court (Judge Ronald L. Buch). Dynamo Holdings Ltd. Partnership v. Commissioner, 143 T.C. 183 (2014), opinion of 17 September 2014. Caselaw Access Project copy of the Tax Court report.

    Mr. Scarazzo opined credibly and without contradiction that petitioners’ approach to responding to respondent’s discovery request is the most reasonable way for petitioners to comply with that request.

Details

AI system

Workflow

The taxpayers restored the backup tapes and removed system files and duplicate messages. The parties were to agree on search criteria for the review. Coding a small sample of documents trains the software to predict relevance.

Follow-up

Conditions for replication

The responding party works with the requesting party's counsel or experts on a search they accept. The requesting party can move to compel if the response proves incomplete.

Entry details

Sector

Legal services / e-discovery.

Purpose

A vendor that the opinion does not name built the predictive coding software to predict which documents answer a discovery request, after learning from a small coded sample.

Not yet checked: model, release, provider, role, configuration, inputs and outputs, autonomy, human oversight, AI contribution, beneficiaries, reach, distribution and access, duration, every measurement question, every money question, ownership and control, dependencies, cross-border connections, geopolitical implications, replication attempts, availability, institutional decision, grounds, action and result, success criterion, every evidence question, observed or intended, alternative assessments, related entries, other identifiers, category mappings.

Cite this record

Peter H. Vartanian. (2026). Tax Court lets taxpayers use predictive coding over IRS objections (AIBD-0833). Artificial Intelligence Benefits Database, edition of 31 October 2026. https://petervartanian.xyz/aibd/records/AIBD-0833.html (archived at https://doi.org/10.7910/DVN/AADSII, version 4.0)

BibTeX Data Field definitions