BCG Applied AI Index 2026: “AI Is Starting to Pay Off”
TL;DR
Boston Consulting Group’s 2026 edition of its AI-maturity survey, now called the Applied AI Index 2026. The full report, The Formula for Agentic AI Value, could not be obtained. This page rests on the press release of 30 September 2026 and excerpts of the publication page.
- 7.5% are “future-built”, “the most AI-mature”. They deliver “2.3 times the total shareholder return, 2.4 times the revenue growth, and 2.8 times the EBITDA growth of laggards” (release).
- 41% are “actively scaling AI”, with “1.8 times the shareholder return of laggards” (release). The publication page defines scaling as companies that “create some value from AI but have the opportunity to do more.”
- “Almost 50%” is the two groups added together: 7.5% + 41% = 48.5%. The release says this “challenges the ‘AI isn’t paying off’ narrative.”
- Spending doubled. AI spending rose “from about 1.7% of revenue in late 2025 to 3.3% now”, and “more than 80%” of it sits outside enterprise IT (release).
- Agents outrun their controls. By 2030, 42% of companies expect to give agents “genuine autonomy”, but “only 5% have the full set of controls in place today” (release). The agentic share of AI value “climbed from 17% in the 2025 sample to 22% in the 2026 sample” (publication page).
Key data
| Figure | Definition | Source |
|---|---|---|
| 1,330 | CxOs and senior leaders surveyed (“more than 1,300 … across 20-plus sectors” on the publication page) | release |
| 7.5% | future-built | release, publication page |
| 41% | scaling | release, publication page |
| 2.3x / 2.4x / 2.8x | future-built vs laggards: TSR / revenue growth / EBITDA growth | release |
| 2.4x | future-built “top-line growth” vs “companies in the bottom half of our sample” | publication page |
| 1.8x | scaling vs laggards: shareholder return | release |
| 1.7% → 3.3% | AI spending as share of revenue, late 2025 → 2026 | release |
| 42% / 5% | expect to grant agents autonomy by 2030 / have the full set of controls now | release |
| ~40% | agentic AI’s expected share of all AI value by 2030 | release |
| 10–15% | expected workforce reduction by 2030 | release |
| 7% → 22% | dedicated AI roles as share of the workforce, 2026 → 2030 | release |
Method and definitions
The release gives the sample (1,330 CxOs and senior leaders) and the group shares. It does not give fieldwork dates, countries, the number of capabilities scored, the score bands or a definition of “laggards”. The 2025 edition used 41 capabilities and four score bands. Whether the 2026 index kept them is not stated in the material read. The release compares future-built firms with “laggards”. The publication page compares them with “companies in the bottom half of our sample”. With 48.5% future-built or scaling, the remaining 51.5% is roughly that bottom half.
Scope and reliability
- Press release only. The full report was not reachable: bcg.com refuses plain curl, the publication page’s download control has no link target, and no PDF was found on BCG’s publication hosts. The publication-page quotes are WebFetch extracts.
- Is it comparable with 2025? BCG does not say so explicitly. Two things point to a continuing series: the release sets 2026 against “Boston Consulting Group’s 2025 research”, and the publication page compares the “2025 sample” with the “2026 sample”. Against that: the study was renamed, the sector count changed from “more than 25” to “20-plus”, and the sample grew from 1,250 to 1,330.
- What the headline compares. The release contrasts 2025’s “just 5%” (future-built only) with 2026’s “almost 50%” (future-built plus scaling). The like-for-like 2025 figure for the combined group is 40% (5% + 35%) in the 2025 report. The outperformance multiples have changed too: 2025 reported 3.6x three-year TSR and 1.7x revenue growth for future-built firms against laggards. The release does not give the 2026 TSR window, so the two sets of multiples cannot be compared.
- Self-report, consultancy instrument. Like the 2025 edition, the maturity grouping rests on executives’ own answers. “Strategic clarity” and “applied AI”, the report’s two dimensions, are also BCG’s corporate positioning: its boilerplate promises “strategic clarity … combined with applied AI.”
Dynamic-capabilities reading
digital-transforming/improving-digital-maturity: the index is a maturity grouping, and dedicated AI roles are expected to triple, “from 7% of the workforce in 2026 to 22% by 2030.”digital-seizing/balancing-digital-portfolios: “61% [of future-built companies] have already moved from isolated pilots to a multiyear, enterprise-wide program”. AI spending doubled to 3.3% of revenue.digital-transforming/redesigning-internal-structures: the release expects “the thinning of coordination and middle-management layers” and a 10–15% workforce reduction by 2030. Future-built firms “make the C-suite responsible for the program.”
How it connects
- BCG’s Widening AI Value Gap (2025): the previous edition, with the 2024 and 2025 group shares.
- enterprise-ai-adoption and ai-maturity-measurement-comparison: the strict figure (7.5%) and the broad one (48.5%) come from the same instrument, depending on which groups are counted.
- agent-oversight-and-delegation: 42% expect to give agents autonomy by 2030, and 5% have the controls.
- Rendtorff-Smith (OECD) on the BCG channel: official firm-level adoption of 20% in 2025, a different instrument and unit.
What was actually ingested
The full press release from BCG’s PR Newswire wire copy (curl), and quotes from the publication page The Formula for Agentic AI Value (WebFetch, cut off by the tool’s length limit). The full report was not read. author: names the firm, because only the press release and an excerpt of the publication page were read. The publication page, read through WebFetch, lists eighteen report authors (below); the release itself names four co-authors (de Bellefonds, Grebe, Walters, Luther).
Linked entities and concepts
- Boston Consulting Group: publisher.
- enterprise-ai-adoption, agent-oversight-and-delegation: concepts.
- ai-maturity-measurement-comparison: thread.
Dangling (deferred): the eighteen authors listed on the publication page. Thirteen of them also sign the 2025 report: Jessica Apotheker, Vinciane Beauchene, Nicolas de Bellefonds, Marc Roman Franke, Michael Grebe, Nina Kataeva, Santeri Kirvelä, Djon Kleine, Romain de Laubier, Vladimir Lukic, Amanda Luther, Jeff Walters, Christoph Schweizer. New in 2026: Dylan Bolden, Sylvain Duranton, David Martin, Natasha Taylor, Rich Lesser.