20 May When Governments Shape Digital Outcomes: What the Evidence Actually Shows
From Singapore’s 18.6% of GDP to Europe’s 8% profit hit — what real outcomes tell digital leaders about policy outcomes for all, including Australia.
We’ve been debating the wrong question
For as long as I have been working in digital technology and observing policy through my company and industry roles, the argument about regulation and innovation has been conducted almost entirely at the level of opinion. Regulation good, regulation bad, government enabling, government blocking, etc.
As digital has become more pervasive across all governments over many years now, and each government pursuing digital policies differently, we can now observe what has actually happened. There is enough evidence across enough jurisdictions, and across enough waves of digital technology, to draw some conclusions that are more grounded than a think-tank position paper. And those conclusions matter directly to every digital leader, digital advisor and policy maker.
We will go through five countries or jurisdictions. Not to place judgment on them but to see what patterns are created, and what those patterns say.
The European Union: heavy regulation but at what cost
The EU has been the most ambitious regulator of digital technology in the world. It gave us GDPR in 2018, the Digital Services Act, the Digital Markets Act, and now the AI Act — the world’s first comprehensive AI legislation, broadly applicable from August 2026. The intent was to build trust, protect individuals, and set a global standard.
The outcomes from these regulations are measurable, and they are mixed at best.
On GDPR, an Oxford University study found that European businesses exposed to the regulation saw profits shrink by an average of 8.1 per cent. The impact on SMEs was sharper: an average 8.5 per cent decline. In the IT sector, small firm profits fell by 12.5 per cent. The number of available apps on the Android platform fell by roughly one third following GDPR’s introduction. Research from multiple economists found measurable reductions in venture capital activity, startup entry, and innovation output. Nothing economically positive in that.
Now the AI Act is following a similar pattern before it has even fully taken effect. The technical compliance standards businesses need won’t be ready until April 2027 — a year after obligations kick in. More than 110 companies, including Airbus, ASML and Mistral, have lobbied for a two-year delay. The EU’s own “Digital Omnibus” proposal from late 2025 is walking parts of it back.
The critics within Europe’s own academic and policy community have noted what the Global Innovation Index confirms. Among individual EU member states, Sweden, the Netherlands, Denmark and Finland rank respectably but as a bloc, the EU has no platform-scale technology company of its own. The complete absence of a European Google, Apple, or Microsoft, despite strong research institutions, is widely considered related to the regulatory environment in which European startups have operated.
As a bloc, the EU’s share of global AI private investment tells a mediocre story. In 2023, EU companies attracted approximately USD $8 billion in venture capital AI investment. The US attracted USD $68 billion in the same year, and in 2024, US private AI investment surged to USD $109 billion while the EU and UK combined remained at roughly USD $9–11 billion. European firms attract just 6 per cent of global AI startup funding, compared to 61 per cent for the US.
As Mario Draghi observed in his 2024 report on European competitiveness, there is no EU company with a market capitalisation over ~USD $100 billion that has been set up from scratch in the last fifty years. By contrast, all six US companies with a valuation above ~USD $1 trillion at the time of his report were created in this same period — a number that has since grown to seven thanks to NVIDIA.
The lesson from Europe is not that “regulation is wrong.” It is that regulation designed without adequate regard for compliance feasibility, the startup economy, or technology maturity creates real, measurable costs. And the burden falls heaviest on those least able to absorb it — the SME and startup businesses. Impacts often over-politely referred to as unintended consequences.
The United States: innovation without guardrails. The results
The US approach has been the inverse. No federal AI law. A light-touch federal posture that has prioritised investment and capability development. The Trump administration doubled down on that in 2025 with a White House AI Action Plan focused explicitly on “winning the race.”
The economic results in terms of innovation output are hard to argue with. In 2024, US private AI investment reached USD $109 billion — nearly 12 times China’s (published) USD $9.3 billion and 24 times the UK’s USD $4.5 billion (Stanford HAI AI Index 2025). The US developed four of the world’s leading AI foundation models — OpenAI (GPT-4/4o), Google (Gemini), Meta (Llama), and Anthropic (Claude). AI adoption across US organisations jumped from 55 per cent to 78 per cent in a single year (Stanford HAI AI Index 2025). The inference cost for a GPT-3.5-level AI system dropped 280-fold between November 2022 and October 2024 (Stanford HAI AI Index 2025).
Those are outstanding results. They reflect what a large, capital-rich innovation ecosystem with minimal regulatory friction can produce. And this has been evident in all my trips there to observe and engage with those ecosystems in 2019, 2022 and 2024.
But the US story also needs to include a state-based overlay. Colorado’s Artificial Intelligence Act, the first US state-level AI law, was passed in 2024 and set to take effect on 1 February 2026. Yet within nine months of its passage, the state’s own Legislative AI Impact Task Force proposed significant amendments — redefining operative terms, reworking exemption lists. Texas introduced a similar bill shortly after. The patchwork has emerged and is creating some instability, although some argue that this is further evidence of healthy competition between states. Perhaps Colorado are advocating for heavier regulation, other states less so.
The concentration of that USD $109 billion in a small number of very large firms — Microsoft, Meta, Alphabet and Amazon together account for the overwhelming majority of AI infrastructure spend, which raises legitimate questions about whether deregulation at scale produces distributed innovation or accelerates concentration.
So, while the macro picture for the US is exemplary, there are some patchiness and risks that have also emerged to satisfy the critics of the low-regulation pathway.
India: public infrastructure as the innovation enabler
India’s story is different in character but also instructive in what it tells us about government’s role in digital innovation. It was a discovery for me in my 2023 visit to India, the discussions with government and many others to understand the incredible role that digital has played in that country’s continuing growth story.
India’s digital economy contributes approximately 11.74 per cent of GDP, around USD $402 billion in 2022-23, and is expected to reach 20 per cent of national income by 2029-30, surpassing both agriculture and manufacturing (State of India’s Digital Economy Report 2024, ICRIER). It employs 14.67 million workers and is nearly five times more productive than the rest of the economy (SIDE Report 2024).
What drove this? Not venture capital. Not light-touch deregulation. It was deliberate government investment in digital public infrastructure:
- Aadhaar — India’s national biometric identity system, launched 2009, now covering 1.4 billion people.
- UPI — the Unified Payments Interface, launched 2016, enabling real-time, interoperable, near-zero-cost digital payments.
- A building stack of interoperable platforms that sit beneath the entire economy.
India now handles 49 per cent of global real-time payment transactions (ACI Worldwide Report 2024). As of mid-2025, UPI is live for merchant payments across more than ten countries including Bhutan, Nepal, Singapore, UAE, France, Sri Lanka, Mauritius, Qatar and Cyprus, with NPCI targeting 20+ countries by FY2029. By May 2025, more than USD $528 billion had been transferred through India’s government Direct Benefit Transfer system since its inception (PIB India, July 2025). Through Aadhaar-based digital verification, over 58.7 million ineligible or duplicate ration cards (government-issued entitlements to subsidised food under India’s Public Distribution System) were cancelled, eliminating systemic welfare leakage at a scale inconceivable without digital infrastructure (PIB India, NITI Aayog).
The regulatory story is more complicated, and not always positive. India’s data localisation requirements and the complexity of the Digital Personal Data Protection Act (enacted 2023, operationalised 2025) are creating cross-border friction that international firms find challenging to navigate (US Trade Representative India Digital Economy Guide, 2026). India ranks 12th among G20 countries in per-user digitalisation depth, strong at the economy-wide scale, still developing at the individual adoption level. Although this latter point is understandable with a population of 1.4 billion spread across urban and rural areas.
The core lesson is clear: government as an infrastructure builder, and not just a regulator, can be a powerful enabler of digital innovation. The private sector built further on top of India’s digital stack.
But government also built in the absence of incumbents. This greenfield environment may be a condition not shared by many others, where vested interests create constraints too complex to navigate.
Singapore: deliberate, proportionate governance
I am always impressed by Singapore, every time I go there. The shared understanding of their strategic challenges and objectives across so many parts of society. And when I was there in 2024 as part of a technology visit that included discussions with Singapore government, that insight was only further confirmed.
If you want to see what it looks like when a government seems to get the balance right, Singapore may be the exemplar.
Singapore’s digital economy reached USD $96 billion in 2024, accounting for 18.6 per cent of GDP (IMDA Singapore Digital Economy Report 2025, October 2025). That is up from 14.9 per cent in 2019, growing at a compound annual rate of 12 per cent over that period, significantly faster than the economy’s overall 7.3 per cent growth rate. Critically, more than two-thirds of that growth came from non-technology sectors: finance, wholesale trade, and manufacturing digitalising their operations.
When it comes to AI adoption, the pace has been remarkable. SME AI adoption tripled in a single year, from 4.2 per cent to 14.5 per cent in 2024. SMEs using AI-enabled solutions under Singapore’s Productivity Solutions Grant reported average cost savings of 52 per cent. Those using AI-powered cybersecurity solutions reported savings of 71 per cent. The tech workforce grew to 214,000 people, with median monthly wages of USD $5,960 per month, well above the overall resident median monthly wage of USD $3,645 per month (IMDA SGDE Report 2025).
What produced this? Not an absence of regulation. Singapore’s Monetary Authority of Singapore (MAS) has been actively regulating fintech, including publishing AI-specific cyber risk guidance, running industry-wide AI risk frameworks through Project MindForge, and maintaining regulatory clarity that the financial industry specifically credits as a driver of its digital expansion (IMDA SGDE Report 2025; MAS Project MindForge 2024). Web3 companies grew from 5 per cent to 16 per cent of the fintech ecosystem between 2022 and 2024, directly attributable to regulatory certainty from MAS.
Singapore’s governance model is deliberate, sector-specific, and outcomes-focused. Rules are set not to constrain innovation but to make it trustworthy enough to scale. The result is a digital economy that now accounts for more than $1 in every $6 of national income (IMDA SGDE Report 2025).
As a result, Singapore is likely the poster child for specific, targeted regulation designed to support rather than stifle growth.
Australia: strong inputs, what happened to outputs
Australia’s GovTech capability is strong, based on the World Bank’s 2025 GovTech Maturity Index, which placed Australia fifth globally with a score of 98.5 per cent, up from 81.1 per cent in 2022. That is a real achievement, driven by investment by government in digital service standards, cloud infrastructure, and interoperability. However, other measures are not so great:
- In the Global Innovation Index, Australia ranks in the mid-twenties, strong on inputs like institutions and human capital, but weak on outputs, such as high-tech exports, resident patent filings, and software exports. Australia ranks 30th on innovation outputs versus 18th on inputs (WIPO GII 2024).
- The weakness is specifically about commercialisation. Australia now ranks 105th out of 145 countries on the Harvard Economic Complexity Index (last among all G20 nations).
- We are a net importer of intellectual property. Imports of IP licences have grown rapidly while our exports, though growing, remain modest (Department of Industry, Science and Resources, Australian Innovation Statistics 2026).
- On venture capital density, Australia lags countries we compare ourselves to, including Israel, UK, USA and Canada. VC investment in Australia is approximately 0.1 per cent of GDP, compared to 0.5 per cent or higher in leading innovation economies (WIPO GII 2025).
We excel in applying digital technology, and genuinely excel, particularly in health, government services, mining, and agriculture, but we have not consistently converted that application capability into exportable, scalable innovation.
On regulation, the December 2025 National AI Plan reflects important priorities: don’t import European compliance paralysis, attract investment, and use existing legal frameworks. Australia’s Productivity Commission rightly warned against regulation as a first resort. The AI Safety Institute was operationalised in early 2026 with AUD $29.9 million in funding (Australian Government, December 2025).
The absence of a dedicated AI law does not mean the absence of regulation. Government agencies are already required to implement formal AI governance under the December 2025 AI in Government Policy. Health organisations operate under TGA frameworks that apply to AI-assisted clinical decisions. Financial services firms work within ASIC and APRA standards that now cover AI use in credit, fraud, and trading. Privacy Act reform, now underway, will mandate disclosure where AI materially influences decisions about individuals.
In Summary
Singapore’s results came from active investment in sectors, active sector-specific regulatory clarity, and direct government support for SME digital adoption. India’s results came from public infrastructure investment that created a platform for private innovation. The US results came from massive private capital deployment. While the European results show the impacts of broad, heavy regulation, especially on the SME and innovation community.
Australia’s evidence also shows that good governance posture alone doesn’t close the commercialisation gap. Australia’s current posture is necessary but not sufficient. Perhaps described as defensive plays, without any offensive plays, if we use a sporting analogy.
For Digital Leaders in Government Policy
- Regulatory burden that outpaces technology maturity and feasibility creates costs, measurable in profits, in startup formation, in innovation output.
- Regulatory clarity that is sector-specific, outcomes-focused, and proportionate enables investment and trust at scale.
- Government as a platform builder, not just a rule-setter, is an underappreciated model, especially where there are gaps in national capabilities.
For Digital Leaders in Industry
- Self-determination remains the powerful remedy to whatever policy environment emerges.
- Building AI governance frameworks, investing in digital capability, and developing internal practices ahead of any mandates do not just check a compliance box, they build the capability that will determine how quickly you can adapt to whatever regulations that do emerge.
- The organisations that treat ambiguity as a reason to wait end up reactive. The ones that treat it as an invitation to lead have some chance to set the standard, including for those standards adopted by the regulator.
For Digital Leaders in Government Delivery
Stay tuned to next month. We will unpack that angle a little more.
Where does your organisation's digital strategy sit against these patterns?
Take the Digital Transformation Strategy Assessment
Mark D Nicholls
CEO and Partner, Information Professionals Group (IPG)
IPG provides independent leadership support for government agencies, utilities, health services, and growing organisations across Australia on ICT and digital strategy, program delivery, and technology transformation.
Know more of our DIGITAL & ICT STRATEGY capabilities. Most organisations are continuously challenged, seeking to stay ahead through
Know more of our DIGITAL & ICT STRATEGY capabilities. Technology in the financial industry would normally bring about thinking
Know more of our DIGITAL & ICT STRATEGY capabilities. A new survey into the scope and frequency of legal
Know more of our DIGITAL & ICT STRATEGY capabilities. Gartner looks into the future, or tries to define it,
- When Governments Shape Digital Outcomes: What the Evidence Actually Shows - 20 May 2026
- The Gartner AI Hype Cycle Reality Check - 17 October 2025
- Digital Project Delivery – How Is It Going, and How to Tell Easily? - 6 September 2024