Highlights from the July 28, 2025 Press Conference: IFF and Deep Knowledge Group Launch Global AI Competitiveness Index, Volume 3 – The Human-Capital Edition

On 28 July 2025 an invited audience of policymakers and press attended the official press conference for the global launch of Volume 3 of the International Finance Forum (IFF) Knowledge Group (DKG) Global AI Competitiveness Index, produced in collaboration with Deep Knowledge Group. The new report pivots the series toward perhaps its most vital pillar: human capital.

View the Full Report Here: https://link.iff.org.cn/AIR3EN 

The launch, held at at Nexxus Building, Central, Hong Kong with virtual attendance by policymakers and press, was presented under the joint patronage of five convening bodies whose presence underscored the importance of the occasion: the International Finance Forum’s Sci-tech Finance Committee, its companion AI Committee, Hong Kong’s own Financial Services Development Council (FSDC), the Office of the Financial Secretary of the Hong Kong SAR, and the Wuhan Industrial Innovation and Development Institute (WHIIID). Each institution brought a distinct vantage-point—capital-markets oversight, technology governance, fiscal policy, and industrial innovation—yet all converged on the same thesis: nurturing and retaining AI talent is now a matter of strategic economic security as vital as access to capital or compute.

The speaker roster reflected that breadth. Le Xia, Chief Economist for Asia and BBVA Research and an IFF Academic Committee member, moderated the afternoon’s flow. Opening keynotes paired public and private-sector voices: King Au (Executive Director, FSDC) and The Honourable Rudolf Scharping (member of the IFF AI Committee and former Minister of Defense of Germany) set the policy stage, arguing that Hong Kong’s regulatory agility can turn it into a global sandbox for AI governance.

Next to speak were Patrick Glauner (Coordinator, IFF AI Committee), Dmitry Kaminskiy (General Partner of Deep Knowledge Group and Member of the IFF Academic Committee) and Zhou Jian (IFF data expert), who walked journalists through the Index’s new human-capital metrics. A later Q&A broadened the dialogue to include Rudolf Mellinghoff (former Justice of the Federal Constitutional Court of Germany) alongside King Au, Kaminskiy and Zhou.

The event marked a milestone not only for the partnership between IFF and DKG, but for Hong Kong’s own ambition to position itself as an Asia-Pacific nerve-centre for responsible, finance-oriented AI development, and the AI Finance Capital of the world. In their remarks the Hong Kong Financial Services Development Council (FSDC), which served as the official observer of the study, stressed that Hong Kong cannot remain a world-class financial hub without being a world-class AI-talent hub, and highlighting how high on the agenda AI already is in Hong Kong, especially in the FinTech and financial services sector.

By bringing finance regulators, veterans of government, academic thought-leaders, data scientists and innovation-zone planners onto one stage, the hosts signalled that talent—not algorithms alone—now sits at the heart of AI power politics. Volume 3’s human-capital focus—and the cross-sector slate that unveiled it—therefore framed the report as both a diagnostic and a call-to-action for economies racing to secure their share of the world’s three-million-strong AI workforce.

As a high-level permanent institution for dialog and multilateral cooperation in the field of global finance established in 2003 by leaders from China, the United States, the European Union, and other G20 nations, along with international organizations such as the United Nations (UN), the World Bank, and the International Monetary Fund (IMF), the IFF plays a crucial role in fostering international collaboration and addressing critical global challenges, and Volume 3 of its Global AI Competitiveness Index serves as an excellent example of its direct and tangible provision of high-quality actionable analytics to key governmental decision makers.

Hong Kong’s Financial Services Development Council: Observer’s Lens and Local Implications

With the Hong Kong Financial Services Development Council (FSDC) serving as an official observer of Volume 3, and the inclusion of its Executive Director as a speaker and panellist, the launch carried an unmistakable local inflection. The Council’s mandate is to keep the city’s financial sector globally competitive, and it has seized on the human-capital focus of the new Index as both a mirror and a roadmap. In opening remarks, the FSDC representative stressed that data-driven visibility on where AI talent clusters—and how quickly they migrates—is now a prerequisite for any jurisdiction that wants to be a magnet for FinTech innovation. Their intervention underlined why the Council chose to observe the project from inception: the Index’s multidimensional indicator system provides the first longitudinal dataset capable of benchmarking how well Hong Kong’s own talent policies and visa schemes stack up against the likes of Singapore, London and New York.

Equally important is the Council’s conviction that finance is morphing into a test-bed for frontier AI. Hong Kong’s virtual-bank regime, maturing digital-asset rules and regulatory sand-boxes mean that applications piloted locally—AI-driven KYC, algorithmic green-bond scoring, predictive credit analytics—can be scaled outward into the Greater Bay Area and beyond. By quantifying the critical mass of engineers, data scientists and academic specialists available in each market, Volume 3 equips FSDC to identify where to source missing skills and how aggressively to price relocation incentives. 

For Deep Knowledge Group and the International Finance Forum the partnership is symbiotic. DKG’s long-standing activity across FinTech, GovTech and InvestTech gives the Index an organic distribution channel into financial-services circles that already trust the consortium’s sector dashboards—an advantage the Council was keen to mobilise from day one.

Finally, the FSDC drew a direct line between the Index and Hong Kong’s aspiration to serve as Asia’s premier responsible-AI finance hub. By treating talent as the foundational layer of competitiveness, Volume 3 offers a scaffold for targeted reforms—expanding postgraduate quotas, subsidising GPU-rich cloud credits for early-stage start-ups, and embedding AI ethics modules in SFC licensing exams. “Without the people, the platforms stay ornamental,” the Council’s delegate concluded, echoing the report’s core thesis. In short, FSDC’s observer role is not ceremonial; it is a prelude to policy actions that could see Hong Kong translate human-capital insights into fast-tracked AI deployment across banking, insurance and capital-markets infrastructure, reinforcing the city’s—and the report’s—mutual stake in the next wave of financial-technology leadership.

Voices from the Podium: Keynotes & Dialogue

Keynotes – “The Future of AI Governance in the World and in Hong Kong” (16:00-16:20): King Au (Executive Director, FSDC) and Rudolf Scharping (IFF AI Committee, former German Defence Minister) opened by framing Hong Kong as a live test-bed for next-generation regulation. Three priorities dominated: developing a cross-border regulatory sandbox built around the Index’s Human-Capital lens to harmonise algorithm transparency and talent mobility across G-20 and BIS jurisdictions; piloting an “AI Compliance Passport” through the HKMA and SFC so firms can trial innovations locally before scaling to the EU or Middle East; and positioning Hong Kong to lead frontier themes such as cross-border model-data hosting, green-compute finance and AI-ethics insurance.

Rudolf Scharping (left) and King Au (right).

Report release – “Global AI Competitiveness Index Insight Report III” (16:20-17:00):Patrick Glauner (Coordinator, IFF AI Committee), Dmitry Kaminskiy (Co-Founder, DKG) and Zhou Jian (IFF Data Expert) presented the new volume as an extension—not a finale—of the series. After noting that Parts I & II are already cited by governments worldwide as policy benchmarks, they detailed how Part III introduces a three-dimensional Talent–Governance–Industry model, quantifies global AI-talent density and mobility, and identifies skills shortages as the chief bottleneck to AI growth. Recommendations ranged from overhauling education pipelines to proactive inclusion of women and minority groups in AI programmes.

Open Q&A (17 : 00-17 : 30): The discussion, moderated by Le Xia (Chief Economist Asia, BBVA), drew in Rudolf Mellinghoff (former Justice, German Constitutional Court) alongside Glauner, Au, Kaminskiy and Zhou. Comments centred on aligning the Human-Capital Index with data-protection jurisprudence, mitigating brain-drain, and balancing algorithmic transparency with national-security considerations.

Rudolf Mellinghoff

Concluding remarks – “Hong Kong’s Role in Shaping AI Governance” (17 : 30-17 : 45): Patrick Glauner and Le Xia closed by reiterating that Hong Kong can leverage its common-law system, capital-market depth and data-port status to export replicable governance models. They singled out three immediate action items: formalising cross-border sandboxes for algorithm validation, operationalising the AI Compliance Passport, and creating an international AI-ethics insurance pool to diversify systemic model risk. Each step, they stressed, draws directly on the talent-centric insights mapped in the new volume, underscoring why human-capital metrics will remain the North Star for the series as it advances toward Volume IV.

Le Xia (left) and Patrick Glauner (right)

Deep Knowledge Group General Partner Dmitry Kaminskiy’s Keynote: Talent Density as the New Geopolitical Fault-Line

Taking the lectern after the FSDC welcome, Dmitry Kaminskiy wasted no time reframing the AI debate around a single variable: people. “AI talent is the most precious asset for all future economies,” he asserted, adding that countries slow to cultivate or import top-tier specialists “risk sliding into a permanent second tier”.

He illustrated the point with a rapid tour of this year’s outliers. South Korea, buoyed by twenty state-funded AI graduate schools and a semiconductor-heavy industrial base, now fields 1 059 AI professionals per million inhabitants—double Mainland China’s density and seven times India’s. That concentration, Kaminskiy argued, is already paying compound dividends: local firms feed on a deep labour pool while global giants relocate R&D to tap visa-fast-track PhD talent.

The keynote then shifted to the Middle East’s cash-fuelled talent surge. In Saudi Arabia, median compensation for senior AI scientists has climbed to USD 420 000 tax-free, with NEOM dangling signing bonuses that “would make even Silicon Valley recruiters blush”; a USD 20 billion pipeline of lab investments aims to create 200 000 new AI jobs by 2030. The neighbouring United Arab Emirates is already reaping early returns, boasting 2 106 AI experts per million residents—out-densifying Germany—and doing so on a workforce barely one-tenth its size. Kaminskiy held these cases up as evidence that aggressive visa regimes, sovereign compute clusters and marquee research grants can “bend the talent curve” within a single index cycle.

Yet absolute scale still matters. Drawing on the report’s macro figures, he reminded the audience that the global AI workforce now sits at roughly three million professionals, seventy per cent of whom live in just five countries. That concentration, he warned, amplifies fragility: sudden immigration-policy shifts or macro shocks in any one of those hubs could upend AI supply chains worldwide faster than chip embargoes ever could.

Kaminskiy concluded by circling back to Hong Kong’s window of opportunity. With financial-services know-how already magnetising fintech-savvy engineers, the city, he argued, needs only “targeted STEM visas, world-class compute and a talent-friendly regulatory sandbox” to punch above its demographic weight. The subtext was clear: the next edition of the Index will likely measure how effectively Hong Kong converts today’s policy talk into tomorrow’s per-capita talent rankings.

1. Talent as Strategic Reserve – Nations ignoring AI human capital “slide into a permanent second tier.”

2. Middle-East Momentum – Saudi Arabia and the UAE have leapt into the global top-20 for AI-talent share through tax-free packages, 10-year Golden Visas and a 36-exaFLOP super-cluster.

3. Density Matters More Than Absolute Size – South Korea’s 1 059 specialists per million people beat several G-7 economies, proving that “talent intensity” predicts innovation velocity. 

4. Tight Labour Markets – Eight of the world’s top-20 economies now report AI-vacancy ratios above 25 %, a signal that supply lags demand. 

5. Hong Kong’s Window – By fusing financial-services know-how with world-class compute and immigration fast-tracks, Hong Kong can “out-punch” larger economies in applied AI finance.

   
   

The Report in Numbers

• 3 million – size of the global AI workforce after analysing >50 000 anonymised professional profiles. (EIN Presswire)

• 70 % – of all specialists reside in just five nations (U.S., Mainland China, India, U.K., Canada). (EIN Presswire)

• 50-fold – spread in AI-talent density across the G-20: Singapore tops at 7 500 per million, Brazil trails at <70. (EIN Presswire)

  
  

Inside Volume 3: Methodology, Metrics, and the Map of Global Talent

When the applause subsided after the keynotes, the spotlight shifted from rhetoric to data. Zhou Jian, lead statistician at the IFF Research Institute, walked the audience through the machinery behind Volume 3. Whereas the first two instalments of the Global AI Competitiveness Index relied primarily on patent filings, venture‐capital flows and benchmark‐test scores, the new report plunges into the human layer—parsing more than 50 000 anonymised professional résumés, university-pipeline datasets and mobility registries to produce the clearest X-ray yet of where the world’s AI brain-power actually resides. This bottom-up census yields a figure that startled even seasoned analysts: roughly three million AI practitioners are active worldwide, but their distribution is wildly uneven. Five countries—the United States, Mainland China, India, the United Kingdom and Canada—claim seventy per cent of that total, with the U.S. and China alone accounting for fifty-seven per cent.

Density metrics reveal an even sharper gradient. By normalising talent counts to population, the researchers uncover a fifty-fold gulf inside the G-20 itself: Singapore tops out above 7 500 specialists per million residents, while Brazil languishes below 70. In the middle sit a handful of fast-mobilising economies—South Korea with 1 059 per million, Saudi Arabia and the UAE now cresting 2 000—whose recent policy sprints show how quickly the curve can bend when immigration, funding and compute all converge.

The dataset also punctures several myths. Only one-third of AI professionals work in pure research or engineering roles; the majority now operate in product, policy or business functions, proof that advanced machine-learning skills are diffusing far beyond R-and-D labs. Eighty-eight per cent hold a master’s or doctorate, yet a striking eighty per cent entered the field after 2022, signalling an unprecedented post-LLM talent surge. Labour-market fluidity, too, turns out to be nation-specific: American specialists have cycled through almost four prior roles on average, versus fewer than three in China—a gap that, as Zhou noted, might explain why U.S. start-ups outrun peers in speed of product iteration. 

Beneath the headline table lies a five-pillar scoring model—technical capability, research output, human capital, policy environment, and infrastructure—each weighted to capture both scale and growth momentum. Volume 3 freezes weights for the first two pillars to ensure comparability with earlier editions, then re-balances the index by inserting three new sub-indicators: talent mobility, graduate-school output and compensation competitiveness. The result is a leaderboard that now shifts markedly from its 2024 configuration: systems-rich but talent-thin economies drop several rungs, while compact nations with aggressive STEM-visa regimes leapfrog into the top ten.

Finally, the authors devote a full chapter to “brain-drain spirals.” By analysing LinkedIn migration timestamps, they trace how incremental outflows from mid-tier talent pools can flip into self-reinforcing exoduses once density falls below a critical threshold. That tipping-point model underpins the report’s blunt policy recommendation: align immigration, compute capacity and cross-disciplinary training—or risk irreversible erosion of national competitiveness. The message landed with force in a room packed with policymakers; the FSDC representative later told reporters that Hong Kong would “use these density benchmarks as our north star for scholarship quotas and investment in AI-finance fusion.

Taken together, the methodological rigour and granularity of Volume 3 transform the Index from a static ranking into what Zhou called a living early-warning system, giving governments and investors a running scorecard on the single most decisive variable in the AI race: people.

Country Playbooks: How Five Very Different Economies Are Bending the Talent Curve

The human-capital lens of Volume 3 shows that there is no single path to AI talent leadership—but there are clear patterns of what works. Below, five case studies illustrate those patterns, each offering practical lessons for policymakers and investors.

Saudi Arabia: “Premium or nothing”: Riyadh has opted for brute-force attraction: tax-free median salaries of about USD 420 000 for senior scientists and signing bonuses at NEOM that climb into seven figures. Layered atop that compensation is a Platinum-Visa regime granting permanent residency in as little as 30 days, plus dedicated special-economic zones designed to soften cultural friction for foreign experts. The Kingdom’s wager is simple—if it can lure 200 000 specialists by 2030, its sovereign-wealth–fund pipeline (USD 1.5 bn dedicated AI fund, 180 000 top-tier GPUs) will find enough human capital to absorb that investment. Volume 3 records early traction: Saudi Arabia now counts 324 AI specialists per million people and has broken into the global top-20 for total talent share. The lesson is that compensation, lifestyle visas and neutral geopolitics can trump geographic size or R&D legacy.

United Arab Emirates: Policy velocity meets compute velocity: Abu Dhabi’s rise rests on three mutually reinforcing levers. First, the world’s first Minister of State for AI (2017) and a policy cadence that delivers new regulations almost yearly. Second, frontier infrastructure: the Condor Galaxy 36-exaFLOP super-cluster, built with Cerebras, gives domestic researchers cloud-class horsepower without U.S. export licences. Third, a talent flywheel: 10-year Golden Visas and AI-fast-track e-visas can be approved in 48 hours, while MBZUAI funnels fresh MSc/PhD cohorts into the local labour pool. The pay-off is visible in the numbers—about 2 106 specialists per million inhabitants, placing the UAE ahead of Germany on a per-capita basis. Speed of policy execution, coupled with sovereign compute, turns a small market into a magnet.

South Korea: Density through academia–industry symbiosis: Seoul treats AI labour as a national industrial policy: twenty state-funded AI graduate schools push out 4 000 advanced degrees annually, feeding chaebol research arms and export-heavy start-ups. The outcome is the highest per-capita AI-talent density in any G-20 economy—about 1 059 specialists per million. Crucially, those specialists sit inside an ecosystem already anchored by semiconductor fabrication, meaning local demand keeps salaries competitive without the brain-drain seen elsewhere. The Korean model underlines how a tight feedback loop—public funding for education tied directly to industry demand—can out-perform larger economies with looser pipelines.

India: Volume today, density tomorrow: India provides seven per cent of the planet’s three-million-strong AI workforce, ranking third globally for absolute numbers. Yet its density—143 specialists per million—remains modest, highlighting vast headroom for upskilling. Two factors give Delhi a strategic edge: an English-speaking diaspora that already supplies 27 % of foreign AI researchers in top U.S. labs, and an IndiaAI Mission (₹10 371 crore) that is wiring college curricula to sovereign LLM projects. If policy can match educational scale with domestic R&D absorption, India could convert numerical advantage into higher-value innovation.

Brazil: The regional first-mover: Measured against its Latin-American peers, Brazil leads on both share and density: 0.5 % of global AI professionals with 69 specialists per million. Brasília’s R $ 23 bn national AI plan underwrites super-computing (the Pégaso 21-petaflop cluster) and open-source Portuguese-language LLMs, serving a vast domestic market starved for locally relevant tools. The takeaway here is strategic specialisation—by owning regional language models and green-energy data-centres, Brazil punches above its weight in attracting talent that might otherwise default to North-American hubs.

Global context: Across all five examples, the macro-picture remains stark: three million AI practitioners worldwide, with a full seventy per cent concentrated in just five countries. The gulf between talent-rich and talent-poor economies thus continues to widen, but Volume 3 shows that deliberate, well-funded strategies—whether salary-driven, visa-driven or education-driven—can bend the curve within a single policy cycle.

Why Human Capital Now?

The first two volumes examined enterprise activity and research output. Yet without skilled people the other pillars collapse. As Prof. Glauner summarised, Talent is the flywheel that keeps the rest of the AI economy turning.

The report’s authors warn of a cascading feedback loop: dense talent clusters attract compute, capital and frontier research grants, which in turn lure yet more experts. For lagging nations, the reverse dynamic—a “brain-drain spiral”—threatens innovation sovereignty.

Implications for Hong Kong and Beyond

A 2023 Cyberport/HKU survey already counts 93,000 finance-tech professionals,  initiatives such as targeted STEM visas, chip-credit subsidies and university-industry AI labs could expand that figure significantly by 2030. By serving as an observer of the index, FSDC gains granular benchmarks that can be used to guide scholarship quotas, immigration quotas and retraining incentives.

Conclusion

The August 28 press conference crystallised a simple but sobering thesis: the race for AI supremacy is, at root, a race for people. That urgency was underscored by the coalition that shared their thoughts on the report at the press conference. 

The International Finance Forum’s Sci-tech Finance Committee and AI Committee supplied the multilateral policy lens; Hong Kong’s Financial Services Development Council and the Office of the Financial Secretary anchored the dialogue in the city’s finance-first development agenda; and the Wuhan Industrial Innovation and Development Institute (WHIIID) bridged the conversation to mainland industrial policy and talent incubation. 

Their joint stewardship of Volume 3 signals that managing the world’s scarce supply of AI specialists is no longer a siloed concern—it is a shared mandate spanning capital markets, public finance and technology-driven industrial strategy.

Volume 3 of the Global AI Competitiveness Index quantifies that race for the first time—and shows that the gap between talent-rich and talent-poor economies is still widening. For policymakers, investors and educators the message is clear: start building, attracting and retaining AI specialists now, or risk watching national competitiveness erode for decades to come.

Download the full report on IFF’s website and follow DKG’s corporate blog for ongoing analysis as Volume 4 moves into production.