A new study from Yijin Hardware has revealed the most and least AI-driven industries of 2025, uncovering a sharp divide in artificial intelligence adoption that experts warn could reshape global economic dynamics.
The study evaluated industries based on a comprehensive AI engagement score, a metric combining global investments in AI start-ups, the number of industry-specific AI ventures, public interest via search trends, and the top use cases of AI in each sector.
According to a spokesperson from Yijin Hardware, the data points to an emerging technological rift across the global economy. “The clear divide in AI adoption across industries reveals a growing innovation inequality that could permanently transform competitive landscapes.”
“This widening gap isn’t just about efficiency gains; it represents a fundamental reimagining of how entire sectors function. Early AI adopters are positioned to completely redefine market rules, while technological laggards face increasingly difficult catch-up challenges,’ it said.
Most AI-driven industries
The most AI-driven industry in 2025 is healthcare, with an AI engagement score of 100, reflecting many AI start-ups and extensive financial support. The global AI start-up investments in healthcare amount to $4.2 billion, including medical AI analysis of X-rays and MRIs, administrative automation, and drug development. Currently, there are 11,228 healthcare-specific active AI startups in the world.
Finance ranks 2nd in the list of the most AI-driven industries in 2025, with a score of 72.4. There is a similar number of AI startups, 11,057, but the funding is smaller. Global investments sum up to $2.1 billion, two times less than for healthcare. The main AI uses in the sector include fraud monitoring and trading automation.
The marketing & advertising sector is third, getting an AI engagement score of 60.9. There is more interest in AI innovations in the sector from the general public than in finance, with 241K monthly searches. The global funding comes to $1.1 billion for over 9,970 AI start-ups.
Legal services follow closely with fourth place and a score of 58.1. The number of AI start-ups in the industry is a little lower than in marketing & advertising, but the finding is much smaller at $0.4 billion, 5 times less than for finance and 10 times less than for healthcare.
Education & learning is fifth in the ranking of the most AI-driven industries in 2025, with an AI engagement score of 41.8. The innovations in this sector are the second most popular, with over 399K queries each month. Most popular AI innovations in the industry are related to virtual tutoring and test grading automation.
Least AI-driven industries
Construction is the least AI-driven industry in 2025, with a minimal possible score of 1. There are fewer than 700 active startups in the sector, with a total funding of $200 million, the lowest in the list.
Insurance takes second place among the least AI-driven industries of 2025, getting an AI engagement score of 5.2. The public interest towards AI innovations for insurance is higher than for construction, with 61K monthly queries, and there are currently 678 active startups in the industry.
Real estate holds third place, scoring 5.4. The funding for it is the same as for insurance AI innovations, at $600 million globally. There are more startups for AI in the sector than in insurance or construction, with 799.
Manufacturing ranks 4th, getting an AI engagement score of 9.5. The general interest in AI in the industry is similar to insurance, but the startups gather more funding, with $900 million. The most explored directions for AI innovations in manufacturing are machine failure detection and quality control.
Agriculture closes the ranking of the least AI-driven industries with fifth place and a score of 11.4. There are over 1,700 start-ups focused specifically on AI in agriculture, developing technologies for spotting insect pests and monitoring crop health. At the same time, the funding for AI in agriculture is smaller than for manufacturing or real estate, at $500 million.
