By Dhruvi Shah
Investments in AI infrastructure and model construction will increase more than five times this year and are expected to exceed $769 billion by 2026, despite AI experts warning of the rapid progress of the technology, as noted in McKinsey’s Technology Trends Outlook 2026.
The projection is based on $384 billion already invested in AI in 2026, based on the trend of the first half of the year. The amount is substantially higher than the $145 billion invested in 2025, making AI infrastructure and model development the largest tech investment area under McKinsey’s investigation.
The spike in investment is due to another round of substantial capital investments of various AI companies. It is reported that OpenAI is negotiating for more funds, potentially bringing its valuation to $1.2 trillion. SoftBank has announced bond sales in the amount of $11 billion aimed at buying a $10 billion stake in OpenAI. Anthropic is also said to negotiate with Nvidia for the possible contribution of $10 billion in relation to the planned IPO worth $100 billion, according to Reuters.
Image:
Despite the rush of finances pouring into the sector, concerns are rising regarding whether innovation is affecting safety measures at the same pace. As a response to the situation, Dario Amodei, CEO of Anthropic, has encouraged the AI field to take a break while safety measures develop. In its turn, OpenAI has been advocating for the establishment of national security standards for advanced AI systems.
OpenAI is also pressing the US to lead international initiatives to establish technical standards for cutting-edge AI systems.
According to the report from McKinsey, there are five sectors of technology that are expected to attract more funds in 2026 than in 2025: Agentic Software Development, AI Infrastructure and Model Designs, AI-driven Scientific Progress, Space Technology, and Robotics. Almost all the trends monitored except for advanced connectivity are expected to show growth.
According to the company McKinsey, this trend is due to the growing significance of the supportive physical infrastructure utilised in the process of scaling AI. This infrastructure involves chips, data centres, and power supply systems. Investment in only the energy technologies amounted to almost $200 billion in 2025, while the expenses on infrastructure for AI doubled in this same year.
Nevertheless, McKinsey states that the upward trend does not necessarily mean that the growth trajectory is bound to continue. Issues like the absence of energy supplies, shortages of skilled workforce and investments, as well as obsolete systems and cybersecurity fears, can considerably slow the development process. In the United States, for instance, AI data centres are expected to consume as much energy as the whole state of California does today. Moreover, there are over 2,500 gigawatts of energy facilities on the waiting line to be connected to the grid.
As regards the business side, McKinsey states that although AI technologies are becoming popular, the organisations are not able to translate the use of the technology into profit. Currently, about 89% of companies are making use of AI in various business processes, and only 37% of them state that the technology is making significant progress.