Sovereign wealth funds power the VC investment wave in GenAI

In 2025, early-stage rounds averaged $28 million, while angel and seed rounds were considerably smaller at just $386,000. This reflects investor preference for larger, more concentrated and perceived lower risk bets.

As in previous years, the highest early-stage valuations are in Series B rounds for pre-establish edcompanies with strong research pedigrees such as Google Research and OpenAI alumni. Recent examples include Distyl AI’s $175 million Series B round for a $1.8 billion valuation⁶ and Together AI’s $305 million Series B round for a $3.3 billion valuation⁷.

Automation is at the core of investment strategy

There is an increasing emphasis on demonstrating return on investment. This is evident in the growing expectation for investee companies to show a clear path to profitability from an early stage.

This heightened demand for robust investment propositions has driven the launch of several new GenAI platforms designed to address well-defined market opportunities.

These include:
OpenAI’s AI-powered browsing tool ChatGPT Atlas that integrates the ChatGPT AI assistant directly into the browsing experience allowing users to access ChatGPT’s capabilities while browsing.

Anthropic’s latest artificial intelligence model, Claude Sonnet 4.5 which offers enhanced abilities in coding while also excelling in specialised fields such as cybersecurity, finance, and research.

Vertesia’s new Autonomous Agent Builder for its enterprise GenAI platform which enables organisations to create and deploy AI agents more quickly and effectively.

VC investors remain focused on GenAI application businesses that build specialised software on third-party foundation models for enterprise and consumer use. There is also growing interest in agentic AI platforms for workflow automation.

Sovereign wealth funds take centrestage in the AI race

In the first nine months of 2025, sovereign wealth funds were involved in AI venture transactions totalling $46 billion. The arrival of these funds on the scene at the same time as the change in the administration in the US is clearly no coincidence. The new administration has earmarked AI as a national strategic priority and has enacted a number of policy changes aimed at preserving US pre-eminence in the technology. These include export controls on semiconductors and other technology and “America’s AI Action Plan,” which is focused on deregulation, innovation acceleration, infrastructure buildout and workforce upskilling.

AI is now viewed as essential for national security and economic growth. Governments are deploying sovereign wealth funds to strengthen domestic AI enterprises and infrastructure, while also making strategic investments in global leaders such as OpenAI.

This new wave of investment is producing some interesting consequences. For example, US export controls are driving investment into Chinese and other alternative semiconductor suppliers. At the same time OpenAI and Anthropic are becoming quasi-sovereign entities through their partnerships with governments.

SWFs, especially in the Middle East, are underwriting massive datacentre buildouts, reducing execution risk and positioning their countries as regional AI hubs. These investments attract private capital and secure access to cutting-edge hardware and foundation models. They embed SWFs in global AI supply chains.

Specific sovereign use cases in areas like defence and healthcare are expected to emerge. In this new era, AI may have the power to shape the outcome of conflicts, making it more logical for nation states to invest in the technology than in missiles. AI will, therefore, be a strategic defence as well as economic asset.

Similarly, at a time when healthcare inflation and ageing populations are threatening to overwhelm health systems and the exchequers that fund them, AI offers a way forward. AI has the potential to deliver new efficiencies, accelerate the development of new therapeutics and revolutionise diagnostics and care delivery.

In seeking to establish their countries as AI hubs, SWFs will not necessarily have to bankroll the entire investment. They will be able to act in anchor roles to underwrite large-scale AI infrastructure supporting the attraction of private capital by reducing execution risk.

North American dominance continues

While the geographic source of the investment capital may have undergone some fundamental changes due to the advent of the SWFs, its destination remains unchanged. Having accounted for 89% of deal value and 70% of deal count in 2024, North America is now responsible for 97% by value and 60% by volume.

That level of dominance is extraordinary by any standard. EMEA activity at around 2% is negligible by comparison, as are the shares commanded by Asia and other markets. While China is competing with the US in AI development, it attracts relatively little external VC funding. This reflects its long-standing preference for a nationally driven approach to financing and innovation.

Looking to the future, both Saudi Arabia and UAE have accelerated their AI infrastructure programmes, while Qatar’s AI market is projected to grow from €380 million in 2024 to €1.8 billion by 2030, a compound annual growth rate of circa. 30%. Across Gulf Cooperation Council member states, GenAI is expected to contribute $23 billion annually, around 2% of GDP, by the same date.

Asia-Pacific is now second only to North America for GenAI adoption. Organisations are moving from pilots to enterprise-scale deployment. For APAC and Oceania, the GenAI market size is estimated at $17 billion for 2025, with growth forecast to be 44.5% CAGR through 2031.

Whilst Google and AWS are committing large-scale AI infrastructure investments in the APAC region, other investors are taking a more cautious stance. They are waiting for clearer regulatory frameworks, proven use cases, and stronger risk management before making significant commitments to GenAI start-ups.

Anchoring Ireland’s AI leadership

With 16 of the world’s top 20 tech multinationals based in the country, Ireland is well positioned to play a pivotal role in advancing AI and driving its integration across industries such as life sciences, financial services, sports tech, manufacturing, and legal services. This advantage is further strengthened by Ireland’s status as the only English-speaking, common-law jurisdiction within the EU that makes it an ideal bridge for US AI companies seeking a European base.

Ireland is not solely dependent on FDI for AI investment. The country is developing its own highly innovative AI sector, with the Ireland Strategic Investment Fund (ISIF) playing an active role.

The ISIF is the State’s sovereign development fund, managed by the National Treasury Management Agency (NTMA). Its remit is to invest on a commercial basis while supporting economic activity and job creation across Ireland.

A key priority for ISIF at present is to invest in scaling home-grown businesses with strong technology and AI capabilities. For example, ISIF has committed €75 million to the Cordiant Digital Infrastructure Equity Fund to drive the expansion of Ireland’s digital infrastructure, focused on data centres, communications towers, and fibre optic networks.

It has also committed €22.25 million to the Molten Ventures Investments (Ireland) L.P. fund⁸ which aims to build a portfolio of early-stage, high-growth technology businesses based in Ireland.

In addition, ISIF has partnered with Elkstone to anchor an early-stage venture capital fund valued at €100 million⁹. The fund’s mandate is to back Irish companies developing innovative, internationally scalable businesses, with a particular emphasis on AI and cybersecurity.

Complementing this, Ireland’s National AI Strategy¹⁰ creates regulatory sandboxes and research hubs, while ISIF provides the capital to scale participating businesses. This is reinforced by Ireland’s strong construction and engineering sector that is recognised for its expertise in building data centre infrastructure critical for AI.

However, an effective moratorium on data centre construction has shifted these firms’ focus to the UK and mainland Europe. In the current climate, it would be somewhat surprising if one or more of them did not attract the attention of SWF investors seeking to bridge skills gaps and develop data centre and other AI related infrastructure in their home markets.

How EU is closing the innovation gap

AI is increasingly regarded as a matter of strategic national importance worldwide, prompting governments to adopt diverse strategies to avoid being left behind. There is, however, ongoing debate about the EU’s approach, with some stakeholders arguing that Europe has prioritised responsible AI and regulation at the expense of innovation.

The EU is responding, and the European Commission launched the AI Continent Action Plan in April 2025¹¹. This is aimed at making Europe a global leader in trustworthy AI by mobilising funds for compute infrastructure, data access, talent development, regulatory simplification, and adoption across strategic sectors.

This was followed in October by the Apply AI Strategy¹², which aims to harness AI’s transformative potential by increasing AI adoption and integration across key strategic sectors such as defence, security, energy, and healthcare.

It also includes targeted measures to boost AI uptake among SMEs. In this regard, the recent proposal to roll back digital protections offers European AI startups some short-term regulatory relief which should enable faster innovation.

However, the Commission’s overall goal is to position the EU as a sovereign, competitive, and ethical AI leader and this means that it must maintain its focus on robust governance.

These are not just policy statements. The Commission has outlined concrete actions: establishing 13 AI factories and up to five gigafactories for large-scale compute and data hubs; mobilising €200 billion through InvestAI plus €20 billion annually for infrastructure and innovation; and training 250,000 AI professionals while expanding EU Skills Academies in AI, quantum, and data science.

While broadly welcomed, OpenAI has urged bolder steps: tripling EU compute capacity by 2030; training 100 million Europeans in AI fundamentals; creating a €1 billion AI Accelerator Fund; introducing a harmonised startup framework by 2026; and launching a Digital Simplification Packageto streamline over 100 tech laws and enable safe experimentation under the EU AI Act.

Building market sustainability

Current anxiety in relation to the prospect of an AI investment bubble should not be treated lightly. The stratospherically high valuations of big tech stocks combined with the eye-watering amounts of venture capital being invested in GenAI companies should be scrutinised when assessing the prospects for AI investments.

History offers cautionary lessons. The dotcom crash of the late 1990s and the unicorn boom before the pandemic, when hundreds of start-ups exceeded $1 billion valuations only to see sharp corrections as inflation rose and interest rates climbed. Few achieved successful IPOs.

If Ireland is to remain attractive for AI start-ups and inward investors, and broader wider tech FDI, it will need to have the compute and data centre capacity available domestically along with other infrastructure.

Currently, developing a data centre in Ireland is highly challenging due to planning restrictions and power grid limitations. For an inward investor, a potential four-year delay would be a significant deterrent, especially when industry estimates suggest an AI gigafactory’s useful life is only around three years. In contrast, locations such as Scandinavia or the UK may appear more attractive under these circumstances. The Kingdom of Saudi Arabia (KSA) too has introduced proposals to attract foreign investment in data centres, allowing international providers to manage hosted data in compliance with foreign legal frameworks.

Similarly, in the energy space, an offshore wind project requires support from vessels which need to be ordered more than a year in advance. Even a slight hiccup at the planning stage could result in a delay of years during the construction phase.

There is, therefore, an urgent need to remove the obstacles and other obstructions which currently exist in the planning and development process. Ireland has already lost out on the European Commission’s AI factories plan and will fall even further behind. Recognising this, the government announced a report and action plan called Accelerating Infrastructure on 3 December 2025, introducing 30 measures to remove barriers.

The strategic importance of Ireland’s maritime area is likely to come into even sharper focus during 2026. Ireland’s energy security is critically dependent on undersea gas and electricity interconnectors, while its tech industry is completely reliant on a network of undersea cables for global connectivity. Recent incidents in the Baltic underscore vulnerabilities. This makes it essential for the Government to review defence arrangements and deploy technology to monitor underwater activity and detect potential threats.