The Hindu Editorial Analysis
04 February 2026
AI’s next investment cycle belongs to applications
(Source – The Hindu, International Edition, Page no.-10 )
Topic: GS Paper – GS 3 : Science & Technology, Artificial Intelligence, Digital Economy, Innovation and Investment
Context
The artificial intelligence (AI) industry has reached an inflection point. After years of massive capital flows into foundational infrastructure — data centres, chips, cloud compute and large language models — the central question is no longer whether AI works, but whether it can generate sustainable profits. The emerging answer is clear: the next phase of value creation in AI will be driven not by infrastructure or training larger models, but by concrete, revenue-generating applications.
AI infrastructure versus AI applications
In 2025 alone, companies globally spent nearly $320 billion on AI infrastructure, yet most foundational model firms continue to post thin or negative margins. High inference costs, intense competition, and pricing pressure have constrained profitability. Even firms reporting strong revenues remain loss-making, relying on venture capital and corporate financing rather than operating income.
By contrast, AI applications tell a different story. Spending on AI applications crossed $19 billion in 2025, accounting for more than half of total generative AI expenditure and over 6% of the global software market — a remarkable achievement barely three years after ChatGPT’s launch. This spending reflects genuine market demand: businesses are no longer experimenting with AI; they are deploying it at scale.
Evidence of real market traction
A growing number of AI products now demonstrate durable revenue streams:
- Over 10 AI products generate more than $1 billion in annual recurring revenue.
- Nearly 50 products exceed $100 million in annual revenue.
- Acquisitions increasingly target application-focused firms rather than infrastructure players.
The acquisition of fast-scaling AI startups illustrates that investors are prioritising tools that deliver measurable business outcomes — automation, productivity gains, and workflow integration — rather than raw computational capability.
Where the real value is emerging
The strongest momentum lies in departmental and enterprise AI applications. In 2025, coding tools alone accounted for $4 billion of a $7.3 billion departmental AI market, making them the largest segment. AI adoption among developers is now mainstream, especially in high-performing firms.
This shift is also reflected in enterprise spending patterns. Application-driven firms are capturing an increasing share of enterprise AI budgets by embedding AI deeply into workflows, leveraging proprietary data, and becoming operationally indispensable. As infrastructure costs fall and efficiency improves, profit margins in applied AI are projected to rise sharply.
Investment patterns confirm the shift
Private equity and venture capital trends reinforce this transition:
- AI application deals increased by 65% year-on-year.
- Nearly 80% of deals were add-on acquisitions for existing portfolios.
- Strategic mergers and acquisitions surged, with deal values rising sharply.
Investors are now seeking companies with real customers, retention metrics, and clear paths to profitability — not just technological novelty.
Core policy and governance challenges
The next phase of AI growth raises complex governance questions:
- Competition concerns as foundational model providers move into application markets, potentially squeezing independent developers.
- Copyright and data ownership issues, as training data and proprietary datasets become commercially and legally sensitive.
- Privacy regulation, especially as AI agents gain access to large volumes of personal and business data.
However, premature or overly rigid regulation could stifle innovation at the application layer. Policymakers must strike a balance — enabling experimentation while enforcing competition safeguards, particularly around acquisitions that eliminate potential rivals through acqui-hiring.
Lessons from past technology cycles
The trajectory mirrors earlier digital revolutions. The Internet was not monetised by selling bandwidth, but by building applications that made bandwidth valuable. Similarly, AI’s commercial success will be determined not by compute capacity alone, but by its ability to solve real problems in healthcare, law, finance, manufacturing, and public services.
Conclusion
The AI sector is transitioning from an infrastructure-heavy, capital-intensive phase to one driven by practical, outcome-oriented applications. Sustainable profits will come from tools that integrate seamlessly into workflows, leverage unique data, and deliver tangible productivity gains.
For investors, governments, and regulators alike, the signal is clear: the future of AI growth lies not in bigger models, but in better applications.