14 technology trends shaping business in 2026

Artificial intelligence is moving beyond screens and software interfaces and becoming part of the physical infrastructure of the economy — from data centres and power grids to robotics, laboratories and manufacturing plants. McKinsey & Company’s latest Technology Trends Outlook 2026 tracks 14 technology trends that are increasingly shaping business and the global economy. Five of them — agentic software development, AI infrastructure and model architectures, AI-enabled scientific discovery and engineering, space technologies and robotics — are on track for investment to more than double in 2026 compared with 2025. At the same time, the next phase of AI scaling will depend not only on better algorithms, but also on energy, computing capacity, semiconductors, infrastructure and skills.

A new technology map

The sixth edition of McKinsey’s report groups 14 technology trends into three broad areas: the AI revolution, new frontiers in computing and connectivity, and advanced engineering. Four trends relate directly to artificial intelligence, but AI is also accelerating development in the other ten, including robotics, semiconductors, cybersecurity, energy, life sciences and biotechnology.

The analysis uses six measurable indicators of activity: internet search interest, media coverage, patents, scientific publications, investment flows and demand for talent. McKinsey supplements those indicators with estimates of adoption levels and interviews with business leaders, investors, policymakers and experts from different industries and regions.

Five trends where investment is on track to more than double
Projected 2026 investment growth versus 2025
Agentic software development>2×
AI infrastructure and model architectures>2×
AI in scientific discovery and engineering>2×
Space technologies>2×
Robotics>2×
Source: McKinsey & Company, Technology Trends Outlook 2026. Full-year 2026 projection based on investment pace in the first half of the year.

Energy is becoming a bottleneck for AI

One of the report’s central conclusions is that AI growth is becoming increasingly dependent on power infrastructure. In the United States, electricity demand linked to AI infrastructure could rise from about 30 GW in 2025 to more than 90 GW in 2030 — comparable with the current power demand of the state of California. Data centres as a whole could account for 14% of US electricity consumption by 2030, up from 3% in 2022.

AI is increasing pressure on the power system
United States: outlook for AI infrastructure and data centres
Power demand from AI infrastructure
2025about 30 GW
2030more than 90 GW
Data-centre share of electricity consumption
20223%
203014%
Source: McKinsey & Company, Technology Trends Outlook 2026.

The challenge is not only the scale of demand but also the speed at which grids can be expanded. More than 2,500 GW of power projects worldwide are waiting for grid connections, while delivery times for critical equipment such as transformers exceed two years in many regions. Capital is also flowing into the energy sector: energy technologies attracted almost $200 billion of investment in 2025, one of the highest totals among the areas covered by McKinsey. Global capital expenditure on data-centre construction could approach $7 trillion by 2030.

“Energy is moving from a secondary consideration to a core constraint on AI infrastructure. The availability of power and the time needed to secure it are becoming structural limits on further growth. That increases the importance of being able to expand power systems quickly and manage them effectively,” said Tomasz Marciniak, Managing Partner of McKinsey & Company in Poland.

Adoption is running ahead of value creation

Rapid adoption of AI does not automatically translate into better financial performance. According to data cited in the report, 89% of organisations use AI regularly and most are at least experimenting with agentic solutions. Yet only 37% say their AI programmes have had any positive impact on EBIT. Implementation costs are also proving difficult to control: 93% of respondents in a McKinsey survey said AI spending exceeded budget assumptions.

AI is widespread, but financial impact remains more limited
Share of respondents reporting selected stages and effects of AI use
Regular AI use in at least one function89%
Higher individual productivity from AI80%
AI scaled across the organisation44%
Positive impact of AI on EBIT37%
Source: McKinsey Global Survey on AI 2026, 1,719 respondents in 97 countries. Data weighted by countries’ share of global GDP.

The gap between adoption and measurable gains is particularly visible in agentic software development. McKinsey estimates that effective deployment could generate almost $1 trillion of value for companies. So far, however, only about one quarter of organisations using agentic coding tools have achieved a meaningful acceleration of the full product-development cycle.

In one study cited in the report, AI increased programming activity by 180%, while the number of software releases rose by only 30%. In 30% of companies, productivity actually declined after agentic tools were introduced.

Chips, cybersecurity and skills for the next phase of AI

As AI models move from training into everyday use, inference — the process in which trained models generate outputs and perform tasks — is becoming more important. McKinsey estimates that by 2030 inference could account for 30-40% of data-centre compute demand and represent most of the compute used by AI systems in day-to-day applications. That raises the value of semiconductors designed for specific workloads, where the goal is higher performance at lower cost and power consumption.

AI is also changing the dynamics of cybersecurity. The same tools that can identify and fix vulnerabilities faster can also increase the speed and effectiveness of cyberattacks. Data cited in the report point to a 44% year-on-year increase in exploitation of vulnerabilities in public-facing applications and a 49% rise in the number of active ransomware groups.

Labour-market data also show that technologies directly linked to AI are still relatively early in their commercial development. Across the four AI trends covered by McKinsey, as well as specialised semiconductors, more than 75% of job postings are for research and development roles. In more mature fields — including cybersecurity, energy, life sciences, mobility and connectivity — more than half of hiring is already for non-R&D functions such as operations, administration, sales and marketing.

The common thread across all 14 trends is growing interdependence. Scaling AI requires parallel changes in infrastructure, operating models, cybersecurity, cost management and skills. McKinsey argues that capturing value from the current innovation cycle will require businesses not only to deploy new tools but also to redesign how organisations operate around them.

“The pace of change creates substantial challenges. Companies are accelerating AI deployment even though proven transformation models and clear scaling paths are still limited. At the same time, they need to build workforce capabilities, modernise legacy systems and strengthen infrastructure resilience against rapidly evolving cyber threats. Access to critical resources — not only energy, but especially new individual skills and new ways of organising work — remains another constraint. Overcoming these barriers requires much more than simply introducing new technologies. It requires end-to-end change management, from operating processes and skills to technology architecture, governance and decision-making,” Marciniak said.

Disclaimer: The information contained in this publication is provided for informational purposes only. It does not constitute financial or any other advice, is general in nature and is not addressed to any specific recipient. Before relying on this information for any purpose, independent advice should be sought. AI may support the editorial team’s work.

Selected categories