VPN Services: AI Market Discovery Index

Benchmark Summary

Proton VPN led the VPN Services benchmark in August 2026 with a valid recommendation coverage of 80.7%, ahead of NordVPN at 77.1% by 3.6 percentage points. The month showed no significant movement across the category, with all ten tracked brands remaining within normal month-to-month variation.

Windscribe recorded the largest coverage increase, rising to 27.6% in August 2026 from 25.9% in July 2026. ExpressVPN saw the largest decline, falling to 67.4% in August 2026 from 71.0% in July 2026. Both moves fall within normal variation for this benchmark.

The benchmark began with 800 prompt-surface observations in each month and produced 707 qualified observations in July 2026 and 709 in August 2026 after qualification.

Current Benchmark at a Glance

Qualified surface breadth counts the canonical AI surface families with at least one qualified observation: ChatGPT, Copilot, Gemini, Perplexity, AI Overviews, and AI Mode. The benchmark captured all six families in both months.

For the strategic interpretation of this benchmark, read CiteWorks Studio's analysis of How AI Search Is Recommending VPN Services

Research Scope and Qualification

The public benchmark is narrower than the raw collection universe by design. The funnel below shows how observations were qualified.

Brand-level percentages use the qualified observations as the public denominator, not the raw collection. See AI Market Discovery Methodology for the full qualification process.

Current Brand Standings

This table is the primary current-month benchmark view, sorted by valid recommendation coverage.

How to Read the Standings

• Presence rate: share of qualified observations where the brand is mentioned in any form.
• Valid recommendation coverage: share of qualified observations where the brand appears in a recommendation context, the benchmark's primary metric.
• Top-three rate: share of qualified observations where the brand appears among the top three recommended options.
• Rank-one rate: share of qualified observations where the brand is the first recommendation.
• Net sentiment: positive minus negative mentions divided by total mentions, scored from -1 to 1.

For formulas and denominator rules, see AI Market Discovery Metric Definitions.

Recommendation Coverage Movement

August 2026 was a quiet month for the VPN Services benchmark. No brand moved beyond normal month-to-month variation, and the category leader, Proton VPN, remained stable. Nine of ten tracked brands saw small declines in valid recommendation coverage, while Windscribe was the sole riser. None of these movements established a trend.

Category Leader: Proton VPN

Proton VPN led the category with 80.7% valid recommendation coverage in August 2026, down from 83.9% in July 2026. Its presence rate rose slightly to 95.8% in August 2026 from 95.3% in July 2026, while its top-three rate was 45.4% and rank-one rate was 14.8% in August 2026. Proton VPN also held the highest net sentiment among the leaders at 0.87 in August 2026, down from 0.90 in July 2026.

Recommendation Placement Snapshot

Coverage alone does not show how prominently a brand is recommended. The leading brands show distinct placement patterns.

Proton VPN held a higher rank-one rate than Surfshark (14.8% versus 6.1% in August 2026) despite a lower top-three rate. NordVPN's rank-one rate of 44.1% in August 2026 shows that similar coverage levels can still hide very different first-position rates.

Buyer-Intent Distribution

All qualified observations in both July and August 2026 fell into the brand recommendation class, which captures discovery and consideration behavior.

The current public series measures brand recommendation discovery only and does not yet contain qualified observations in the pricing and value or multi-brand comparison classes.

Historical Measurement Record

Evidence and Source Layer

The benchmark is built from prompt-level observations that retain the query, AI/search surface, answer, brand outcome, recommendation placement, sentiment, and, where exposed, citations or attributable evidence sources. Source presence is evidence about the information environment. It is not automatically proof that the source caused the recommendation.

Scope Boundaries

This public benchmark does not measure:

• Market share or attributable sales
• Every possible AI response across all model versions and configurations
• Organic-search ranking or search engine optimization performance
• Social media mention volume or sentiment outside AI surfaces
• Private or sponsored channels, including gated enterprise deployments
• Causality from a metric movement alone; movements indicate correlation, not cause

About This Benchmark

The LLM Authority Index AI Market Discovery Index is a neutral, industry-level benchmark tracking how AI and search surfaces recommend brands across canonical surface families. It is built from prompt-level observations with audited, reproducible methodology.

• AI Market Discovery Methodology
• AI Market Discovery Metric Definitions
• AI Market Discovery Research Standards
• Modeled AI Authority Value

Get a Company-Level Authority Index

The public industry benchmark shows category-level standings. A company-level Authority Index can go deeper.

The public percentage cannot identify the prompts, competitors, or sources causing the result.

Get my free AI Company Index