- AI is becoming an enabler of significant cost savings and revenue generation across industries, fuelled by massive global investments in infrastructure and technology
- Its continued development will create winners and losers – for investors, identifying resilient, adaptable companies capable of integrating AI is essential
- We believe the coming decade promises unprecedented innovation – with the potential for long-term investment returns – if approached strategically and responsibly
By BNP Paribas Asset Management Chief Market Strategist, Daniel Morris, and co-Head, Environmental Strategies Group, Edward Lees
Artificial intelligence has moved from experimentation to large-scale deployment and a tangible revenue driver across an ever-broadening range of industries.
As AI models become more powerful more companies and sectors are seeing increased adoption and emerging value and profit growth. And we believe that AI has the potential to become pervasive throughout many different sectors of the economy.
For example, logistics companies can leverage AI for route optimisation, with potentially large cost savings. Farmers use AI-driven systems for targeted herbicide application, cutting operating costs while also reducing environmental pollution.
Utilities can employ AI for predictive maintenance, reducing downtime, while renewable projects use AI to analyse weather patterns to maximise yields. And healthcare companies can harness AI in developing new drugs, aiding discoveries and reducing time to market for potentially life-saving treatments.
Soaring capex requirements
But AI development requires substantial investment and infrastructure. Recent estimates suggest the top US hyperscalers – massive cloud operators – are set to spend some $785 billion on AI and infrastructure this year, and close to $1 trillion in 2027.1
Further research predicts nearly $3 trillion of AI-related infrastructure investment globally by 2028.2
The return on investment will depend on AI’s ability to deliver measurable value – and while momentum is building, returns will take time to emerge across industries.
Some 85% of organisations increased their AI investment over the previous 12 months, according to a 2025 survey by consultancy Deloitte, albeit with most respondents achieving a satisfactory return on investment within two to four years – longer than the typical seven to 12 months expected for technology investments.3
For the hyperscalers themselves, the picture is more uncertain. Significant capex in building out AI infrastructure requiring everything from power and chips needs to be matched with attractive investment returns in the hope that lower cost of usage will increase adoption over time.
Winners and losers
AI adoption will undoubtedly create winners and losers, with companies embracing AI’s capabilities potentially able to improve efficiency, grow their business and improve performance, with their slower-to-adapt rivals being left behind.
There will also be leaders and laggards within the technology sector. Whilst software was first to fall victim to fears of AI disruption, we believe that software providers that develop systems vital to a business’s core operations – known as ‘mission critical’ software – are likely to thrive as AI enhances their offerings.
Conversely, legacy software companies may face obstacles and even obsolescence if they fail to adapt to lower barrier to entry.
Presently, hardware firms, especially those supplying semiconductors, are benefiting from soaring demand. But in some cases, margins may start to come under pressure. For investors, identifying resilient, adaptable companies capable of integrating AI into their core strategies is essential.
The expansion of AI infrastructure is also driving a surge in power demand, but even more cheap and readily available renewable energy is vital to sustain this growth, while battery storage and grid innovations are equally crucial. Companies in these areas, as well as those providing components for energy infrastructure, are benefiting from high long-term demand, creating potential new investment opportunities.
The drive for AI sovereignty
As nations recognise AI’s strategic importance, debates about AI sovereignty are becoming more prominent. Countries seek control over their own data, hardware, and software – which could potentially create fragmentation across a global tech market into regional blocks that in turn could increase costs and slow innovation.
For investors, navigating this geopolitical landscape requires understanding regional policies and the potential for cross-border collaborations. We believe that companies with global footprints and flexible strategies may be better positioned to thrive amid evolving regulations.
AI’s automation of tasks also raises concerns about job displacement. However, we believe that long-term technological disruption can also create new roles and industries. At the same time, AI can augment human capabilities, enabling workers to focus on higher-value tasks.
Policymakers and AI companies must collaborate to promote reskilling, transparency, and responsible deployment, minimising societal disruption and ensuring that people as well as companies benefit from AI’s potential.
The future of AI
We are already seeing AI evolve from chat-based systems to autonomous agents capable of completing complex tasks independently. As this shift accelerates, we believe it will drive breakthroughs in areas ranging from scientific research to physical AI applications like robotics.
As AI systems become more agentic and capable of self-improvement, the pace of innovation will intensify. However, this underscores the importance of responsible AI governance: balancing technological progress with energy efficiency, security, and ethical considerations.
AI is already moving towards tangible profitability, driven by massive investments and groundbreaking applications across sectors. For investors, we see potential opportunities in identifying resilient companies, understanding the evolving landscape, and navigating geopolitical nuances.
The coming decade promises unprecedented innovation, if approached strategically, with the potential for long-term investment returns.
[1] Moody’s: Hyperscaler capex forecasts marked up by $85bn, to close in on $1trn by 2027 Data Centre Dynamics
[2] AI Market Trends 2026: Global Investment, Risks, and Buildout | Morgan Stanley
[3] AI ROI: The paradox of rising investment and elusive returns
