1 of 22 independent experts
Voting member of the expert panel, representing Romania as founder and CEO of TVL Tech, alongside experts from the OECD, the World Bank, Stanford University, and Harvard Kennedy School.
WITSA Global AI Index 2026 · Independent Expert Panel
I helped validate the WITSA Global AI Index 2026.
I was one of 22 independent experts on the panel behind the first WITSA Global AI Index, a diagnosis of how 128 economies have actually built their AI ecosystems. I sat on it as founder and CEO of TVL Tech, representing Romania, and helped review and sign off the methodology behind every score.
- 128 economies
- 7 pillars
- 83 indicators
- OECD/JRC compliant
The appointment
A seat on the panel, representing Romania.
Authority over the methodology
Under the index Terms of Reference, the panel signed off the indicator list, the sub-pillar framework, the pillar weights, and the methodology document itself. The experts could change the methodology, and they did.
128 economies, 105 scored
Seven pillars, 16 sub-pillars, and 83 indicators, built to the OECD/JRC Handbook on Constructing Composite Indicators.
Published in 2026
The inaugural edition of the Global AI Index from WITSA, the World Innovation, Technology and Services Alliance, a global alliance of technology industry associations.
What it measures
Ecosystems that were built, not announced.
The index scores AI ecosystem development, the structural conditions that decide whether AI produces durable national value, not the announcements or the strategy documents behind it. Scores are per-capita by design, so a small country building intensely is not buried by a large one, and each pillar runs 0 to 100 against the observed global frontier for that measure.
Infrastructure
Compute, connectivity, and the physical layer AI runs on.
Talent
Researchers, developers, and the people who build with AI.
Entrepreneurship
Startups, unicorns, and the pace of company formation.
Use Cases
Real adoption of AI inside organizations.
Data
Open data, datasets, and published models.
Policy
Governance, strategy, and AI readiness of the state.
Investments
Funding, greenfield capital, and financial commitment.
Why it is credible
Open sources, and open to challenge.
An independent host
WITSA carries no vendor interest and no government affiliation, which is the argument for letting it hold the measuring instrument.
Documented sources
Every indicator has a documented source, and every imputed value is flagged rather than hidden.
Contestable data
Every economy can contest a specific indicator value through a formal data review process run by the Secretariat.
A panel with authority
The full methodology was reviewed by 22 independent experts who had the authority to push back before they would validate it.
Romania's scorecard
Strong on data and skills, thin on capital.
48.0overall EDS
Data87.7
Talent59.9
Policy55.4
Use Cases42.3
Infrastructure33.9
Investments31.4
Entrepreneurship25.6
Source: WITSA Global AI Index 2026, confirmed pillar scores (final frozen run). Pillars run 0 to 100 on a per-capita basis.
Romania's clear strength is Data, at 87.7, one of the higher data scores in Europe, followed by Talent at 59.9 and Policy at 55.4. The constraints sit on the commercial and physical side: Entrepreneurship at 25.6, Investments at 31.4, and Infrastructure at 33.9. This is an economy with strong human capital and a strong data foundation whose AI ecosystem is held back by weak capital formation and a thin physical layer. For scale, the index is led by the United States at 82.2, Singapore at 82.0, and Sweden at 79.9. Luxembourg, where I chair the AI Committee of the Romania Luxembourg Business Forum, scores 59.2 overall, led by Talent at 83.2.
Why it matters
A national score is made of company decisions.
Romania's weak spots, turning capability and capital into deployed use cases, are exactly what I see inside individual firms as AI that stalls after the pilot. The panel measures where the country is weakest; my consulting and training work is aimed at the same pillar, Use Cases, one company at a time, getting teams past experiments and into production systems they actually use.
The same reading, applied to your organization.
For AI strategy, enterprise implementation, or a briefing on the index and what it means for your teams, email me directly with your objective and timeline.
alex@tvl.tech