Global AI Adoption Statistics: A Review from 2017 to 2025

The speed with which AI entered our lives is phenomenal. 

It changed most people's perception of artificial intelligence. I remember seeing AI as a technology that simply delivers an output, such as a suggestion or action, based on the input. Like when a piece of software detects a customer’s frustration in their voice and flags them as a top priority. 

Many companies did something similar while marketing themselves as AI-powered in the initial phase of AI adoption. But the real potential came with generative AI. 

This might be a surprise, but long before ChatGPT and DALL·E became household names, some enterprises were already using AI in 2017.

This article covers AI adoption statistics from 2017 to 2025. It shows the yearly trends and how we entered the current AI hype. 

AI adoption: Evolution of intelligent technology at a glance

Sources: McKinsey: State of AI in 2024, McKinsey: AI in the Workplace, Statista, Reuters, and NY Times

AI adoption from 2017 to 2025: 9 years in review

Below are a few statistics that showcase how AI has evolved over the past nine years. 

2017: 20% of enterprises adopted AI

In 2017, only 20% of survey respondents confirmed they adopted AI in at least one business area. 

McKinsey’s State of AI in 2022 covered AI evolution between 2017 and 2022. The survey observed that even though AI adoption was 2.5x higher in 2022 than in 2017, it increased to 58% in 2019 but dropped to 50% in 2022. 

Besides growth, here are a few additional events that happened around AI adoption in 2017: 

• The number of AI papers published each year increased nine times compared to the 1996 data. 
• The number of active US startups developing AI systems increased 14 times since 2000. In 2017, around 600 startups developed AI systems. The annual VC investment increased by 6x in the same period.

• Job openings that needed AI skills in the US increased by 4.5x since 2013.
• Error rates in image labeling fell from 28.5% in 2010 to 2.5% in 2017.
• An AI system trained on a dataset of 129,450 clinical images of 2,032 diseases could classify skin cancer at a level of competence comparable to that of a dermatologist.
• In 2017, the proportion of corporate AI papers in the U.S. was 6.6x that of corporate AI papers in China.

2018: AI Adoption grew by more than 100%. 

The McKinsey report estimated that AI adoption was 47% in 2018, compared to only 20% in 2017. This is more than double the adoption rate in 2017, showing more than a 100% increase in adoption. 

On the academic side, AI was a trending topic. The 2018 Advancement of Artificial Intelligence (AAAI) conference was held in February in New Orleans, Louisiana. The conference observed that 70% of papers submitted were affiliated with the U.S or China. However, the number of papers accepted was remarkably even, 268 and 265, respectively.

Below are some more interesting events that happened around AI adoption in 2018. 

• The U.S.-affiliated papers had an acceptance rate of 29% vs. China-affiliated research papers at 21%. 
• Attendance at the International Conference on Learning Representations (ICLR) 2018 grew 20x since 2012. The conference focused on deep reinforcement learning within AI. 

• WiML alone saw 750+ women participate — a 600% increase from 2014.
• Active AI startups in the U.S. increased 113% between 2015 and 2018.
• Articles on AI became 1.5x more positive from 2016 to 2018

Based on these statistics, AI adoption trended upward in 2018. Companies like Amazon and Alphabet invested $16.1B and $13.9B in research and development related to AI, respectively.

2019: AI adoption in organizations reached 58%

Considering the ongoing development and research in AI, 2019 showed an increase in AI’s adoption rate. Besides this, many events happened in the AI space during this period, including:

• The share of AI-related job postings in the U.S. rose from 0.26% in 2010 to 1.32% by October 2019. Machine learning topped the charts and made up 0.51% of all job listings.
• Washington had the highest share of AI job postings at 1.4%, followed by California and Massachusetts at 1.3%, New York at 1.2%, DC at 1.1%, and Virginia at 1%.

• In 2019, private global investment in AI crossed $70 billion. 
• Startups raised over $37 billion, and mergers and acquisitions (M&A) were valued at $34 billion. IPOs brought in $5 billion, and minority stakes attracted around $2 billion.

From mid-2018 to mid-2019, over 3,600 global news articles explored AI-related topics. They were primarily focused on fairness, interpretability, and explainability.

2020: AI adoption in large enterprises fell to 50%

COVID-19 clearly showed its impact on AI adoption in 2020. With priorities shifting and budgets tightening, the adoption rate fell to 50% in 2020. Still, not everything slowed down.

• Investment in AI for drug design and cancer research hit $13.8 billion in 2020, 4.5x higher than in 2019. This was the top-funded AI sector globally.
• In 2020, technologies like facial recognition, video analytics, and voice ID became more accurate, affordable, and widespread. It drove broader surveillance capabilities worldwide.
• Although many groups produced ethical frameworks and principles, a consistent way to measure or evaluate AI development was absent. 
• In 2020, around 4.2 billion digital voice assistants were predicted to be used worldwide. 
• The U.S. posted 8.2% fewer AI jobs in 2020 than in 2019. The jobs dropped from 325,724 to 300,999.

Although overall adoption was comparatively slower, some sectors, like healthcare, observed high investment in AI-related R&D. 

2021: AI adoption climbed back to 56% in organizations

Some survey respondents suggested their AI investments didn’t increase despite the global COVID-19 pandemic. Participation in in-person events moved to an online medium instead. Here’s a snapshot of a few interesting events in the AI space in 2021. 

• AI journal publications grew by 34.5% from 2019 to the start of 2021. 
• Major AI conferences shifted online due to COVID-19. As a result, the attendance doubled across nine major events.
• Generative AI made substantial progress. AI was able to create realistic text, audio, and images. 

• AI came closer to human performance in language tasks. On basic reading benchmarks (e.g., SuperGLUE), AI was able to beat humans by 1–5%. For complex tasks like abductive natural language inference (aNLI), the human-AI performance gap shrank from nine points in 2019 to just one in 2021.
• Robotic arms became more affordable. The median price dropped by 46.2% in five years, from $42,000 in 2017 to $22,600 in 2021. This made robotics research more accessible.
• AI investment surged to $93.5B in 2021, more than double 2020 levels.

Generative AI rose in 2021, and many big organizations increased their investments. Private investment in AI reached approximately $93.5 billion in 2021, more than twice the amount invested in 2020. However, the number of newly funded AI startups declined from 1,051 in 2019 to 762 in 2020 and further down to 746 in 2021.

2022: AI adoption plateaued at 50%

After bouncing back, enterprise AI adoption dipped again to 50% in 2022, indicating that the industry had reached a plateau. However, this trend was similar to other technologies in the early years of their adoption. 

Michael Chui, Partner at McKinsey Global Institute, said, “We might be seeing the reality sinking in at some organizations of the level of organizational change it takes to embed this technology successfully.” 

Companies that thought implementing AI would be a quick exercise were discouraged, but those that grew their AI muscle slowly incorporated more AI capabilities. 

Here’s an overview of everything that influenced AI adoption in some way in 2022:

• The U.S.–China collaboration in AI research saw the most cross-country activity from 2010 to 2021. It increased by 5x over the decade despite rising geopolitical tensions. 
• Companies used more AI tools. On average, 3.8 different AI capabilities in 2022, up from 1.9 in 2018.
• The most common use of AI remained the same for four years: companies leveraged AI to optimize service operations.

• 54% of the world’s large language and multimodal models in 2022 came from U.S. institutions.
• In 2022, the industry dominated AI development. Industry produced 32 major models, compared to just three from academia.
• The year 2022 saw the public launch of generative tools like DALL·E 2, Stable Diffusion, ChatGPT, and Make-A-Video.
• For the first time in a decade, global AI private investment declined 26.7%, from $93.5B in 2021 to $91.9B in 2022. Still, AI funding in 2022 was 18x higher than in 2013.
• 37 countries passed AI laws in 2022. This number increased from 1 company in 2016.  
• The U.S. remained the global leader in AI investment, and attracted $47.4B in 2022, 3.5 times more than China.

Although investment decreased and adoption plateaued, AI models, laws, policies, and capabilities rose in 2022. This was likely when the organization gained maturity in implementing AI, leading the industry toward the AI spring. 

2023: AI adoption grew to 55%

In 2023, AI adoption in organizations increased to 55%, while 33% of survey respondents confirmed that their firm used generative AI in some way. 

• In 2023, industry-led AI research released 51 key machine learning models, while academia contributed only 15. Collaborations between industry and academia hit a record with 21 joint models.
• AI organizations released 149 foundation models in 2023 —  more than double the number in 2022. Nearly 66% of these models were open-source, up from 44.4% in 2022.
• U.S.-based institutions produced 61 top AI models in 2023, more than the EU (21) and China (15) combined.
• China led the world in AI patent origin with 61.1% in 2022. The U.S. followed with 20.9%, down from 54.1% in 2010.
• Generative AI investment skyrocketed to $25.2B in 2023, 8x more than in 2022.

• The U.S. AI investment hit $67.2B in 2023, nearly nine times more than China. China and EU investments fell sharply.
• The U.S. AI regulations grew by 25 in 2023, up from just one in 2016.
• 21 U.S. agencies regulated AI in 2023, up from 17 in 2022.

Investments in generative AI increased in 2023, creating the initial setup that fueled its rise in 2024 and 2025. Let’s look at what exactly happened in 2024. 

2024: 72% of organizations used AI

From 50% adoption in 2022, AI adoption surged to between 72% and 78% in 2024, depending on which study you trust more. 

Personally, I feel the adoption rate is a little higher than what’s been reported. Stanford’s 2025 AI Index reported, “78% of organizations used AI in 2024.” Either way, it’s clear that the AI adoption trended upward. McKinsey and Stanford’s data reflected this. 

Here are a few relevant statistics to this adoption trend: 

• In 2024, the industry built almost 90% of the top AI models, up from 60% in 2023. Academia led in the most highly cited AI papers.
• The U.S. produced 40 top AI models in 2024, ahead of China (15) and Europe (3)
• Video generation from text improved in 2024. New tools like OpenAI’s SORA and DeepMind’s Veo 2 create better video content.
• AI incident reports hit a record with 233 cases in 2024, up 56% from 2023.

• 8 out of 10 professionals prioritized AI capabilities when selecting software.
• 40% of companies relied on automation to streamline and improve data entry.
• 83% of organizations that purchased an AI solution saw a positive ROI.
• 75% of businesses implemented two to five AI features, suggesting a measured yet committed approach. Meanwhile, 17% have adopted six to eight features across their operations.
• Only 2% of organizations reported quick AI adoption in IT. However, marketing emerged as the fastest department to adopt AI, according ot 53% of organizations. 
• Only 26% of companies turned AI pilots into real business value. Moreover, only 4% were at the cutting edge of AI maturity. 
• 62% of AI value came from core business areas like operations, sales, and R&D.
• Only 10% of AI implementation challenges came from AI algorithms, yet many companies wrongly overfocused here. 70% of obstacles were people- and process-related.
• BCG reported that without decisive action, 75% of companies risked falling behind in the AI race.
• Mentions of AI in global legislative records rose 21% in 2024.
• The U.S. federal agencies introduced 59 AI regulations in 2024, up from 25 in 2023.
• AI optimism rose globally, from 52% in 2022 to 55% in 2024.
• 60% of people believed AI would change their job; only 36% feared it would replace them.

AI adoption reached an all-time high, with a rate between 72% and 78% globally. With continued investment, this rate is expected to rise even further in 2025. 

2025: Entering the intelligent age 

McKinsey research suggests that almost all companies are investing in AI, but only 1% believe they’re at maturity. The challenge is not employees but leaders who are not moving fast. While companies are looking forward to the long-term gains of AI, 92% are planning to increase their investment in the next three years. 

Below is an overview of what’s latest and yet to come in the AI space in 2025. 

• 69% of C-suite companies began investing in generative AI over a year ago. Despite that, only 47% say they are making slow progress in building Gen AI tools.

• C-suite leaders are 2.4x more likely to say employee readiness is a significant barrier to adopting AI. But employees are using generative AI three times more than their leaders think.
• 48% of employees rank training as the most crucial factor for gen AI adoption.

These stats and other ongoing trends suggest that AI isn’t plateauing anymore; it’s on the rise from an innovation, implementation, and adoption perspective. 

An MIT Technology Review article suggests that large language models (LLMs) will be able to “reason.” It talks about 2023 being the age of generative images and predicts 2025 to be the age of generative virtual playgrounds. 

Sector snapshots: Key AI adoption figures across industries

Below are concise stats of AI adoption by industry, illustrating how AI use differs across industries: