Tech Investments Face New Challenges Amid AI Growth

Corporate software spending is declining, and software revenue as a percentage of US GDP is easing.

[SINGAPORE] Investors who once enjoyed substantial returns from software companies are now facing uncertainty as the world increasingly adopts artificial intelligence (AI).

According to a report by Bain & Co, the era of easily profiting from software-as-a-service investments is ending. Despite tech deals comprising 22% of all buyouts in North America in early 2025, up from 19% at the end of 2024, the market penetration in many areas is flattening. Software spending and its contribution to US GDP are also decreasing.

Software companies are no longer exploring untapped markets but are competing in areas already utilizing software. Bain notes that while sectors like construction may still have potential, overall growth from entering underserved markets will be harder to achieve.

Future returns will depend on finding new revenue sources and improving operational efficiencies through competition displacement, AI utilization, modern pricing models, and market expansion.

Software firms can also integrate payment capabilities or monetize data for new revenue streams. AI's demand for compute power now exceeds Moore's Law, with computational needs growing at more than twice the rate over the past decade.

Bain forecasts global compute requirements could reach 200 gigawatts by 2030, with the US accounting for about half. "AI compute demand is outpacing semiconductor efficiency, necessitating significant power supply increases on grids unchanged for decades," said David Crawford, Bain's global technology practice chairman.

To address this, algorithms must improve to reduce computational load, and technological breakthroughs like specific AI chips over general-purpose GPUs are needed for better power efficiency.

"Navigating the potential for innovation, infrastructure, supply shortages, and algorithmic gains is crucial for the coming years," Crawford added.