McKinsey’s Newest Analysts Are AI Agents

The article assesses how AGI could pressure professional-services firms that bill by the hour, while favoring companies able to monetize AI through access, outcomes, or higher-priced subscription tiers. RELX and Intapp score 4 of 4 on the article’s defensibility test; RELX reported 10% underlying legal growth in the first half, with three-quarters of renewal value on its higher-priced Protégé AI tier, while Intapp revenue grew 15% and cloud net revenue retention was 123%. Other signals are mixed: Accenture and ExlService each score 1 of 4, Gartner and Robert Half score 0 of 4, and Am Law 100 firms’ 2026 aggregate revenue rose 13%; RELX’s Oct. 22 update and Thomson Reuters’ Nov. 3 report are identified as upcoming catalysts.

Analysis

The investable distinction is not “AI exposure” but who captures productivity savings. Workflow and proprietary-content vendors can potentially shift value from labor hours to higher-priced access; labor-based services risk giving those savings to clients unless they can reprice around outcomes. That makes RELX’s renewal monetization more decision-useful than broad claims of AI adoption, while Intapp’s leverage could prove durable if agents remain embedded in its system of record. Neither moat is absolute: model providers can pressure content pricing, and clients may use agents to reduce seats or challenge database dependence.

Near term, RELX’s update on Oct. 22 is the key test; Thomson Reuters’ Nov. 3 disclosure is a second catalyst. Look for evidence of realized price/mix and renewal retention, not just AI-tier adoption. Over 1–3 months, weak monetization evidence could prompt multiple compression across legal-information names; over 6–18 months, labor substitution may pressure consulting, outsourcing and staffing economics, but current weakness in Gartner and Robert Half is not proof of AI causality. Macro demand and valuation could dominate both directions. The article’s scorecard is a useful screening heuristic, not a valuation framework: strong operating metrics may already be priced in, and a small sample of disclosures does not establish durable pricing power.

Key Decisions for Investors

• Event watch: consider a measured RELX position only if the Oct. 22 update sustains legal growth and confirms customers are paying more at renewal for AI-enabled access. Avoid chasing a sharp gap; trim or stand aside if growth decelerates or pricing/retention evidence is absent.
• Keep Thomson Reuters as a relative-value watch, not an automatic short: the Nov. 3 catalyst is explicit disclosure of AI-tier price realization and renewal behavior. A failure to show either while RELX demonstrates it would support a relative long RELX versus Thomson Reuters, subject to entry valuation and borrow/liquidity checks.
• Do not short Accenture or ExlService solely on the labor-hour thesis. Track contract mix, revenue per employee, and realized client pricing; improving output-based economics would falsify the bearish read, while continued revenue growth without per-unit monetization would strengthen it.
• Risk control: verify current valuation, renewal cohorts, and post-announcement price response before sizing. A deterioration in RELX legal growth, weaker retention, or evidence that clients can substitute general-purpose AI for proprietary research would invalidate the preferred quality tilt.