Karnov Q2 2026 slides: AI workflow launch drives margin expansion

Karnov’s Q2 2026 results and AI-driven workflow launch sparked a 12.86% pre-market stock surge to $81.6. The company reported 7% Online Legal growth (from 6%) and expanded adjusted EBITA margin to 25% (from 23%), alongside leverage improving to 2.0x (from 2.5x), though organic growth slowed to 2% (from 5%) due to weak France offline demand. Karnov also highlighted pricing power with the AI layer lifting the subscription value index to at least 130 (vs 100 historically) and suggested a potential workflow tier could push the index to 200+, while continuing SEK 159m of share repurchases and keeping capex at 6–8% of net sales.

Analysis

The market is likely pricing this as proof that vertical, proprietary data plus workflow AI can justify both higher ARPU and a higher terminal multiple. The real implication is not just better monetization for one vendor; it widens the moat gap versus generic legal-tech tools that depend on public statutes and LLM wrappers, which can’t easily match defendable outputs or procurement-friendly audit trails. That is constructive for scaled content owners like RELX and TRI over 6-18 months, while smaller AI-point solutions without owned content should see multiple pressure if investors start demanding proof of retention and pricing power.

The second-order loser is offline/training-heavy revenue streams and any business line tied to discretionary legal education spend, because AI workflow adoption shifts budget toward subscription software and away from lower-margin services. The weak French book is important: it shows the transition is uneven and that legacy channels can be structurally eroded even when the core platform improves, which argues against extrapolating the headline growth rate too aggressively into the whole enterprise. If the workflow module does not show measurable conversion in the next 1-2 quarters, the rerating can fade quickly because the current move already discounts a strategic premium.

Contrarian view: consensus may be overestimating the speed of enterprise conversion and underestimating procurement friction in public sector and large-firm accounts. The best falsifier is simple: if recurring net adds or price uplift stall into autumn 2026 despite feature releases, the AI narrative becomes incremental rather than transformational. For GS, the M&A-advisory angle is positive but too small to matter for the stock; the opportunity is in reading this as a sector signal, not a standalone bank catalyst.

Key Decisions for Investors

• Long RELX / TRI on any AI-commoditization dip over the next 1-3 months: this release supports the thesis that owned content monetizes AI, and these are the cleaner ways to express it with lower execution risk than a small-cap single name.
• Avoid chasing the post-print rally in proprietary-content small caps until autumn 2026 workflow adoption data is visible; use a 1-2 quarter watchlist trigger on conversion, retention, and price uplift rather than the headline margin expansion.
• Short a basket of public-data dependent legal-tech / document-automation names if borrow is available; the edge is that they lack the defensible content moat and should see multiple compression as investors pay for audited, defensible outputs.
• No core trade in GS; any M&A fee is de minimis. Only consider a tactical long if the strategic review broadens into a formal process with dates, bids, or breakup value disclosure.