That makes it one of the most AI-optimizable practice areas in law.

Antitrust law runs on data — market definitions, concentration ratios, pricing analysis, competitive effects modeling. That makes it one of the most AI-optimizable practice areas in law. The question isn't whether AI helps with antitrust work. It's whether your firm can compete without it when the opposing team is using Harvey's antitrust agents and you're still having associates manually compile market share data.

The landscape shifted dramatically when Allen & Overy (now A&O Shearman) partnered with Harvey to build antitrust-specific AI agents. These tools analyze merger filings, identify competitive concerns, and generate substantive analysis at a speed that traditional methods can't match. The DOJ and FTC are also increasing enforcement velocity — more second requests, more merger challenges, tighter timelines. AI isn't a luxury in antitrust practice anymore. It's a capacity requirement.

Market Analysis and Definition: AI's Core Antitrust Application

Defining the relevant market is the foundation of every antitrust case — and it's where AI adds the most value.

The traditional approach: Economists spend weeks compiling market data, analyzing substitution patterns, calculating HHI concentrations, and defining geographic markets. This involves public data (industry reports, SEC filings, Census data), proprietary data (client sales records, pricing data), and expert judgment.

The AI approach: Claude analyzes industry reports, public financial filings, and market research to generate preliminary market definitions in hours. Feed it SEC 10-K filings from the merging parties and their competitors, and it identifies product overlaps, geographic reach, customer segments, and competitive dynamics. This isn't the final market definition — that still requires economic analysis — but it creates a working framework that would take an associate 2-3 days to compile manually.

Harvey's antitrust agents (enterprise, A&O Shearman partnership) go further — they're trained on antitrust-specific data and can analyze merger filings against historical enforcement patterns. Harvey identifies which product markets are likely to raise concerns, which geographic markets overlap, and which competitive dynamics the agencies will scrutinize.

HHI calculations: The Herfindahl-Hirschman Index is simple math but requires comprehensive market share data. AI collects and organizes the data from public sources faster than manual research. Claude calculates pre- and post-merger HHI levels and identifies which markets exceed the DOJ/FTC merger guidelines thresholds (2,500 HHI for "highly concentrated").

Caution: Market definition is ultimately an economic exercise that courts treat as a question of fact. AI-generated preliminary analysis is a starting point for your economists, not a substitute for expert testimony.

Merger Review and HSR Filing Support

Hart-Scott-Rodino filings generate massive document review obligations — the second request is essentially a comprehensive discovery process run by the government. AI transforms this from a nightmare into a manageable project.

HSR filing preparation: The initial HSR filing requires industry codes, revenue data, competitive overlap descriptions, and prior acquisition information. AI organizes this information from client records and public filings. A filing that takes a team 2-3 days to prepare can be assembled in 4-6 hours with AI assistance.

Second request compliance: When the DOJ or FTC issues a second request, the merging parties must produce all documents responsive to broad specifications — often millions of documents. This is e-discovery at its most intensive. Relativity and Everlaw handle the document review platform; AI-powered TAR identifies responsive documents from the massive collection.

But the antitrust-specific challenge is different from ordinary e-discovery. Second request specifications ask for documents relating to competition, market definition, pricing strategy, and competitive effects — concepts that require antitrust expertise to code properly. AI review models trained on antitrust document sets (available in Relativity and Everlaw) achieve better accuracy than general-purpose TAR models because they understand antitrust-specific language.

Timing pressure: Second requests typically must be substantially complied with in 60-90 days. For a company with millions of potentially responsive documents, AI-assisted review is the only way to meet the deadline without hiring an army of contract attorneys. Firms that handle second requests without AI review technology are either over-staffing (expensive) or under-producing (risky).

Bloomberg Law's antitrust resources track merger enforcement trends, helping predict which deals will face challenges. This intelligence informs the go/no-go advice that clients pay premium rates for.

DOJ/FTC Filing and Regulatory Response

Engaging with the antitrust agencies — DOJ Antitrust Division and FTC Bureau of Competition — requires a specific type of legal analysis that AI accelerates.

White papers and position statements: When clients need to persuade the agencies that a merger is pro-competitive or that their conduct doesn't violate Section 1 or Section 2, the supporting analysis is data-intensive. Claude drafts initial position papers by analyzing the market data, identifying efficiency justifications, and organizing competitive effects arguments. A 30-page white paper that takes a partner and two associates a week to draft can be first-drafted in 2 days with AI assistance.

Consent decree analysis: When the agencies propose consent decrees or behavioral remedies, AI analyzes the proposed remedies against historical consent decrees in comparable cases. Claude can identify whether proposed remedies are typical, unusually aggressive, or potentially insufficient — intelligence that informs the client's negotiation position.

CID response management: Civil Investigative Demands (CIDs) from the FTC are effectively government subpoenas requiring production of documents and information. AI assists with responsive document identification, privilege review, and response drafting — the same document review workflow as second requests but in an enforcement context.

Historical enforcement analysis: AI's ability to analyze patterns across hundreds of enforcement actions is genuinely valuable. "How has the FTC treated vertical mergers in healthcare in the last 5 years?" Claude answers this question with specific case citations and outcomes in 15 minutes. An associate doing the same research manually needs 4-6 hours.

Cartel Detection and Criminal Antitrust

Criminal antitrust enforcement — price-fixing, bid-rigging, market allocation — involves pattern detection across communications and pricing data. AI excels at finding the patterns that prove (or disprove) coordinated behavior.

Communication analysis: In a cartel investigation, the key evidence is often buried in thousands of emails, text messages, and meeting notes. AI review tools identify communications between competitors that discuss pricing, market allocation, or bid coordination. The AI flags language patterns associated with anticompetitive coordination — references to "the agreement," competitor pricing discussions, production limitation conversations.

Pricing pattern analysis: Claude can analyze pricing data across competitors and time periods to identify patterns consistent with price-fixing — parallel pricing movements, price increases that lack independent justification, or suspiciously similar bid submissions. This statistical analysis supports or undermines the existence of an agreement.

Leniency program analysis: Companies considering applying for DOJ leniency (amnesty) in a cartel investigation need rapid assessment of their exposure. AI reviews the company's communications and pricing data to assess the strength of the evidence and the likely scope of the conspiracy — critical information for the leniency application decision that must be made quickly.

Internal investigation support: When a company discovers potential antitrust violations internally, the investigation must be fast and thorough. AI-assisted document review identifies the scope of the problematic conduct, the individuals involved, and the time period affected. Speed matters — in criminal antitrust, the first company to apply for leniency gets amnesty. The second gets prosecuted.

Building an Antitrust Practice with AI Capabilities

Antitrust practice has traditionally been concentrated in AmLaw 50 firms because of the resource requirements — large teams, expensive economists, months-long document reviews. AI changes the economics enough that mid-size firms can compete for antitrust work.

The resource shift: A mid-size firm with AI tools can handle merger analysis, HSR filings, and civil antitrust litigation that previously required 15-attorney teams. AI replaces the junior associate research hours, not the senior partner strategy. A team of 5 attorneys with AI tools can handle the workload of a 12-attorney team without AI.

Tool stack for antitrust practice: - Harvey (if budget allows) for antitrust-specific AI analysis - Claude for market analysis, filing preparation, and research - Relativity or Everlaw for second request and document review - Bloomberg Law for antitrust enforcement tracking and regulatory intelligence - Lex Machina for antitrust litigation analytics (judge behavior, case outcomes, damages data)

Revenue potential: Antitrust work commands premium rates — $600-1,200/hour at major firms. Mid-size firms can compete at $400-700/hour while maintaining strong margins because AI reduces the associate hours required. A single merger review engagement can generate $200,000-2,000,000 in fees.

The competitive positioning: Position your firm as "the antitrust team that uses AI to deliver faster analysis at lower cost." Clients — especially mid-market companies facing their first antitrust issue — respond to efficiency and cost predictability. They don't need an AmLaw 10 firm for every antitrust matter. They need competent counsel with modern tools.

The Bottom Line: Harvey for firms that can afford enterprise AI with antitrust-specific training. Claude + Bloomberg Law for mid-size firms building antitrust capabilities at lower cost. Relativity for second request document review — it's non-negotiable for serious merger work.