Artificial Intelligence's Real-World Impact

Over the past few years, artificial intelligence (AI) has moved through three eras: training, where models learned from vast datasets; reasoning, where models could think through multi-step problems and support complex decision-making; and agentic AI, where systems act autonomously to achieve defined goals.

The latest era is the most defining step-change yet, driven by significant leaps in model capabilities that shift AI from a tool that only responds to one that plans, executes, and refines tasks on its own. This is what makes ‘code creating code’ not just a possibility but an operational reality.

Even two years ago, investors were still questioning whether AI could create real economic value. That is no longer the case, with the shift from concept to real-world results already underway. We are seeing this across our portfolio:

SOFTWARE

In software, large language models are now widely used to support coding tasks across the industry. For example, Anthropic’s agentic coding tool, Claude Code, optimises software development by acting as an autonomous coding agent that can navigate, edit, and test code across complex, end-to-end workflows. Claude Code applies leading engineering practices and identifies performance bottlenecks faster than traditional manual development, helping developers accelerate delivery and improve code quality.

FINANCIAL SERVICES

In financial services, AI is accelerating product innovation and enabling client-led AI workflows. For example, financial operations platform Ramp has used these capabilities to rapidly deploy features such as custom roles and expansive integrations, driving significant enterprise traction and supporting geographic and currency expansion. By embedding autonomous internal agents and integrating tools that can connect smoothly with other systems, Ramp is positioning itself as a leader in using AI to automate and improve financial operations.

HEALTHCARE

In healthcare, leading technology provider athenahealth is adopting AI to improve patient care, automate clinical workflows, and reduce administrative burden on medical practices. This includes AI-powered appointment scheduling and AI scribes for medical documentation and diagnosis. athenahealth is also integrating AI in its reimbursement processes, reducing administrative tasks and enhancing productivity. For example, compared to manual processes, some medical practices recovered 30% more revenue from insurance claims that were initially denied by following athenahealth’s AI medical coding advice.

These examples show how quickly AI capabilities are generating business impact. AI adoption is increasingly prioritised.

In our Q1 2026 survey of nearly 200 North American enterprises across technology, healthcare, financial services, consumer, and other sectors, more than half of the respondents ranked AI among their top three priorities. Within this group, one in five placed it as their topmost priority. The survey also found that quality factors, such as performance, accuracy, and data security, drove foundation model selection more than costs. Importantly, enterprises were also reaping savings from AI adoption in a wide range of domains such as finance, customer service, customer onboarding, and non-software research and development (see Figure 1).

Figure 1. Reported Cost Reduction from AI Adoption

Cost reductions can be seen across a wide range of domains beyond coding, and at varying levels.

Significant Cost Reduction (By >50%)

Moderate Cost Reduction (By 20-50%)

Modest Cost Reduction (By <20%)

Unsure or No Cost Reduction

Share of respondents

Source: GIC Enterprise AI Survey, Q1 2026

These early signs of real value present clear opportunities for investors. Yet, as AI scales, the landscape is also becoming more complex to navigate.

Challenges of Investing Amidst a Fast-Evolving AI Landscape

Even as AI’s economic impact becomes clearer, the fast-evolving AI landscape raises challenges for investors to identify long-term value:

Pace of model innovation

AI models are advancing at an exponential pace. Today, they can reason through multi-step problems, maintain context across longer interactions, use external tools autonomously, and orchestrate complex workflows. New interoperability standards, such as the Model Context Protocol, are also making it easier for AI systems to link with external tools and with one another. However, near-term breakthroughs are not always indicators of the capabilities or companies that will dominate over the long term. Competition is intense, with AI researchers racing to improve model capability, cost, and reliability. Competition is also playing out across countries, with distinct AI ecosystems emerging based on their strengths and constraints in computing power, advanced semiconductors, and energy. Many governments are treating AI as a strategic priority, creating domestic demand and support for local champions. For investors, the key question is which parts of the technology stack will retain pricing power as the frontier evolves.

Capital intensity of AI infrastructure

Rapid progress in model capability is being matched by an equally sharp rise in compute demand, driving hyperscaler capital expenditure to record levels—now approaching 2% of United States GDP. Training and serving frontier models require vast amounts of computing power, memory, networking capacity, and electricity. Bottlenecks are already emerging even as there are efficiency gains from better design and custom chips. This will help cope with growth in demand, but AI infrastructure's supply chain and power constraints will persist. This creates opportunities for investors to fund critical infrastructure but also makes the path ahead less linear and the long-term outlook less clear.

Uneven adoption across industries

Even as AI infrastructure and model capacity scale rapidly, adoption remains uneven. Full adoption across industries will take time as enterprises have to manage security, compliance, and data governance. A clear divide between AI leaders and laggards will emerge as enterprises adopt AI at varying pace. At the same time, new business models will emerge to bridge this gap. For example, Anthropic, Blackstone, and Hellman & Friedman's new AI services company aims to help midsized companies deploy AI by embedding small engineering teams to design and implement AI workflows. Such partnerships can help more enterprises turn AI progress into practical, secure, and scalable business solutions. Investors will need to differentiate between companies that can turn AI adoption into compounding advantage and those which are at risk of being disrupted.

Navigating these challenges requires a disciplined framework guiding where to look and what to look for to find companies with lasting value.

Finding Lasting Value: Where to Look and What to Look For

We use two frameworks to navigate the AI landscape: one that maps where to look across the AI value chain, and one that defines what to look for in individual companies.

Where to Look

The first framework helps assess opportunities across the AI value chain, splitting the investment universe into three distinct categories.

Enablers

Companies that build the foundational infrastructure that powers AI, including the computing hardware layer, high-speed networks, and energy systems;

Monetisers

Companies that develop AI-powered products and services; and

Adopters

Companies that integrate AI into their existing operations to improve processes, boost productivity, and unlock growth.

We apply this framework across public and private markets since leaders in each category exist in both. For example, in enablers, the hardware layer has been driven largely by public-listed semiconductor leaders, whereas the development of large language models has seen private companies take the lead. While this framework effectively guides us on where to look for opportunities, more aspects need to be considered to identify companies with lasting value.

What to Look For

The second framework provides us with a bottom-up view to understand which company is well-positioned to succeed over time in each category.

Moat

We look for business models with moats, or structural advantages, that can thrive as AI reshapes industries. Examples of moats include: proprietary data; strong governance around critical workflows, particularly where there is little margin for error; and products and services that are resilient to AI-driven disruption.

Management

Successful companies will require management with both vision and execution capability. Strong leaders understand how AI drives real business outcomes and formulate strong AI strategies, rather than only rolling out AI tools. They also attract and retain the talent to execute these strategies. The most forward-looking leaders use AI not just to cut costs but to strengthen core capabilities. Those who invest in reskilling and preparing their workforce for new ways of working are more likely to build organisational resilience and lasting value.

Momentum

As AI evolves rapidly, successful companies can build momentum that reinforces itself. This is because early wins generate better data, which leads to better AI performance. Companies that are able to harness this cycle across industries are likely to gain lasting advantages.

The GIC Advantage: Broad Platform, Deep Expertise

GIC is among the few global investors that have both early- and late-stage investment teams and the ability to invest across public and private markets. This breadth gives us a complete picture of the rapid innovation with frontier models, capital flowing into AI infrastructure, and longer-term adoption trends across industries and geographies.

Since the 1980s, GIC has been participating in early-stage investing and has stayed through multiple market cycles. This longevity matters not only for generating good long-term returns for our venture portfolio but also for building depth of insight and strong partnerships. Being involved early helps us see which technologies are taking off and how they are being used in practice, which founders can execute well, and which business models have lasting advantages. GIC’s partnerships with leading venture and growth funds also give us access to a broad range of companies which we can start early engagement with.

In public markets, we built early conviction in AI through close engagement with hyperscalers and semiconductor leaders. The most critical advances in foundational computing infrastructure, networking, and memory are concentrated in a small number of large, listed companies. Strong relationships with these companies have helped us stay ahead of the curve.

With a network that spans both incumbents and start-ups across public and private markets, GIC is well-positioned in the technology ecosystem. We engage with this ecosystem in many ways.

Our Technology Investment Group (TIG), headquartered in San Francisco, invests in leading technology companies globally and is deeply embedded in key innovation hubs across the US and in India. Formed in 2017, the team has grown to around 20 investment professionals focused on AI, software, fintech, and more. As a lifecycle investor, TIG deploys capital across venture, growth, and public markets, enabling GIC to invest in companies from growth to scale.

Within GIC, our Technology Business Group brings together our technology investors across asset classes, and in both public and private markets. This enables us to share insights, develop a unified view of the rapidly evolving technology landscape, collaborate on strategic investments, and direct capital to where it is most needed.

GIC also connects stakeholders in the technology ecosystem through flagship events including: Bridge Forum, which convenes global business leaders and technology trailblazers to connect, share exclusive insights, and develop investment opportunities; Partnership Forum, which brings together portfolio company founders, executives, industry advisors, and investment teams to discuss opportunities for value creation; and GIC Insights, our annual thought leadership event that gathers global business leaders and policymakers to discuss long-term issues. We also work with our investee companies and partners to host targeted sessions that facilitate the exchange of insights among founders, fund managers, and industry peers, strengthening connections across the technology and start-up community.

AI’s trajectory is clear, but its path is nonlinear. GIC's ability to invest early and across companies’ lifecycles, as well as our deep engagement with the global technology ecosystem gives us a unique vantage point to invest well in this space.

Beyond the Hype: Investing in Artificial Intelligence Value