What real usage data says about AI adoption in business: who uses it most, where the biggest gains come from, and what smaller teams should do next.
Every week, dozens of articles appear claiming that AI will “change everything.” The problem is simple: most of them are based on forecasts, surveys, and marketing claims. Not real data.
In March 2026, Anthropic published the latest edition of its Economic Index – a large-scale study based on 2 million real conversations with AI. Not “we asked 500 managers,” but actual data on who uses artificial intelligence, how, and for what, right now.
In this article, I will break down the key findings, show what they mean for Ukrainian businesses, and give you a concrete action plan: which processes can already be automated and where to start.
Key findings from the Anthropic Economic Index
Anthropic Economic Index is not a marketing report. It is a research project that analyzes real AI usage through an economic lens. The study covers conversations on Claude.ai (consumer use) and through the API (business integrations).
Here is what the latest 2026 data shows:
AI use is becoming more diverse
The top 10 most common tasks accounted for 24% of all conversations in November 2025. In February 2026, that number dropped to 19%. In other words, people are finding new and new uses for AI, moving beyond “write text” and “help me code”.
Experienced users get better results
People who have used Claude for more than 6 months have a 10% higher success rate on their tasks compared with beginners. This does not depend on task type, country, or model version. The simple takeaway: AI is a skill, and it improves with practice.
Augmentation has overtaken automation
52% of Claude.ai conversations are augmentation (the person works together with AI in a dialogue). 45% are automation (AI performs the task on its own). But in API integrations the picture is reversed: 75% of traffic is full automation.
What does that mean? Businesses that integrate AI through APIs (chatbots, automated funnels, scripts) are already operating at an automation level. Individual users, by contrast, mostly use AI as a smart assistant.
AI cuts task time by an average of 80%
An analysis of 100,000 conversations showed that tasks that used to take about 90 minutes now take an average of 18 minutes with AI – an 80% reduction. More complex tasks get even bigger gains: college-level tasks speed up 12x, and high-school-level tasks speed up 9x.
49% of professions now use AI for at least a quarter of their tasks
In January 2025, 36% of professions had at least 25% of tasks supported by Claude. By the end of 2025, that share rose to 49%. AI is not limited to IT – it is gradually spreading across more and more industries.
Who uses AI the most, and for what
The Anthropic data makes it clear who is currently leading AI adoption:
Programmers and technical specialists – 35% of all conversations
Computer and mathematical tasks account for a third of all Claude.ai conversations and almost half of API traffic. Developers write code, debug, create documentation, and test. These are the first and most active AI users.
Marketers and copywriters
Content generation, ad copy, email newsletters, and social posts. Marketing is one of the fields with the highest AI adoption: 76% of marketing teams in small and medium-sized businesses already use AI.
Management and operations roles
Reports, analytics, data organization, and presentation prep. Directive delegation (when the user gives AI more autonomy) rose from 27% to 39% in 8 months – people are increasingly trusting AI to carry out complex tasks on its own.
Sales, HR, and legal teams – the next wave
These areas are not leading yet, but usage is growing quickly. According to BigSur AI, 83% of growing SMBs have already adopted AI in their processes.
An important trend: the gap between large companies and small businesses in AI adoption is shrinking. In February 2024, large businesses used AI 1.8x more actively than small ones. By August 2025, that gap had almost disappeared: 10.5% versus 8.8%.
What this means for Ukrainian businesses
Anthropic Economic Index data shows a clear correlation: a 1% increase in GDP per capita equals a 0.7% increase in AI adoption. Wealthier countries use AI more intensively.
For Ukraine, this means two things at once:
Challenge: the overall AI adoption level in the country is lower than the leaders (the US, Japan, and the UK). Microsoftโs AI Diffusion Report notes that Ukraine has strong technical talent, but strategic investment in AI skills is still not enough.
Opportunity: small businesses can adapt AI faster than corporations. While big companies negotiate budgets and build compliance processes, a sole proprietor can connect a Telegram chatbot in one evening and start automating customer support.
That is why AI is a window of opportunity for Ukrainian small business:
- Low barrier to entry. ChatGPT Plus costs $20/month. That is less than lunch at a restaurant. And there are even more free tools.
- Scale compensation. One entrepreneur with AI can do the work of a small team: write copy, analyze data, generate creatives, and reply to customers.
- Technical literacy. Ukrainians are among the most active freelancers on Upwork. The technical base for AI adoption is already there.
- Decision speed. Small businesses do not need six months for a pilot project. Try it, see if it works, scale it. If not, try something else.
The key point: AI is not a “technology for big companies.” Anthropic Economic Index shows that the biggest productivity gains are happening exactly in knowledge work inside small teams. If you work with your brain, AI can already help you.
5 processes AI is already automating
No theory – just concrete processes businesses are automating right now, with tools and real time savings.
1. Content creation
What: website copy, social posts, product descriptions, blog articles, and video scripts.
Tools: ChatGPT, Claude, Jasper, Copy.ai.
Time saved: copy that used to take 3-4 hours now takes 30-60 minutes with AI (draft + editing). That does not mean “AI writes it for you” – it means you spend your time on strategy and editing instead of staring at a blank screen.
2. Customer support (chatbots)
What: answers to common questions, product guidance, lead collection, and routing inquiries.
Tools: custom chatbots built on OpenAI API, Tidio, Intercom, ManyChat, and Telegram bots.
Time saved: chatbots handle 60-80% of routine inquiries without human involvement. For a small business, that can replace a support hire – or let the owner stop answering the same question 20 times a day.
3. Analytics and reporting
What: analysis of data from GA4, Google Ads, and CRM; report creation; insight extraction from large datasets.
Tools: Claude (CSV/PDF uploads), ChatGPT Advanced Data Analysis, Looker Studio + AI, Perplexity for market research.
Time saved: analysis that used to take 4-6 hours (data gathering, processing, conclusions) now takes 1-2 hours with AI. Claude can process 100+ pages of a report in a minute and surface the key insights.
4. Ad copy generation
What: headlines for Google Ads, Shopping descriptions, Facebook/Instagram ad copy, and A/B variants.
Tools: ChatGPT, Claude, Google Ads AI features, Performance Max.
Time saved: 15 RSA headlines in 5 minutes instead of an hour of brainstorming. Plus, AI generates variations you would not think of on your own, which improves A/B testing.
5. Email marketing
What: newsletter copy, segmentation, personalization, and trigger sequences.
Tools: ChatGPT + any email platform (MailChimp, SendPulse, Brevo), Zapier for trigger automation.
Time saved: creating a full 5-email sequence used to take 2 days. With AI, it takes 3-4 hours. Personalization by audience segment is even easier because AI adapts tone and message to different groups quickly.
AI diagnostics for business: 15 processes to automate
PDF: check which processes in your business can already be automated with AI. 15 processes, tools, and expected time savings. Save it and act on it.
Get it on Telegram →How to start: 3 levels of AI maturity for business
Not every business is ready for the same level of AI integration. That is normal. Figure out where you are now and move gradually.
Level 1: Manual AI use
What it is: you open ChatGPT or Claude and ask questions manually. You ask it to write copy, analyze something, or generate ideas.
For whom: anyone just starting out. Sole proprietors, freelancers, microbusinesses.
Tools: ChatGPT (free or $20/month), Claude (free or $20/month), Perplexity.
Result: 5-10 hours saved per week on routine tasks. Better content, faster replies, less research time.
How to move up: once you see which tasks repeat every day, it is time to automate them.
Level 2: Integrated AI tools
What it is: AI is built into your workflows. Zapier handles inquiries automatically. The CRM suggests the next step. An email flow triggers without your involvement.
For whom: businesses with 1-10 employees that already have stable processes and want to scale them.
Tools: Zapier ($20/month), Make.com, HubSpot AI, Notion AI, automated Google Ads strategies.
Result: processes work without you. You focus on strategy, not operations. 15-20 hours saved per week.
How to move up: when standard integrations are not enough and you need solutions tailored to your business.
Level 3: Custom AI solutions
What it is: an AI chatbot trained on your data. An automated analytics dashboard. An AI agent that performs complex tasks on its own.
For whom: businesses that already passed levels 1-2 and want a competitive advantage through technology.
Tools: OpenAI API, Anthropic API, custom Telegram bots, AI agents, and custom automations.
Result: AI becomes part of the product or service. Customers get a better experience. The business scales without proportional team growth.
Anthropic data confirms this model: Claude.ai is mostly augmentation (levels 1-2), while the API is mostly automation (level 3). Businesses that reach level 3 get the biggest acceleration.
Want to define your AI maturity level and build an implementation plan?
I can help diagnose your business and choose tools that fit your tasks and budget. Message me on Telegram ยท My services
Common questions about AI in business
According to Anthropic, yes on average. But it is an average. Simple tasks (rewriting text, finding information) speed up 5-10x. Complex tasks (data analysis, strategic planning) speed up 2-3x. The key factor is prompt quality: experienced users get 10% better results than beginners.
Start with ChatGPT Plus ($20/month) – it covers 70-80% of needs: copy, ideas, analysis, code. If you work with long documents or need precision, add Claude. For market research, use Perplexity. For process automation, use Zapier. Do not buy everything at once: one tool per month, assess the result, then add the next one.
Anthropic data shows that 52% of AI use is augmentation (humans and AI working together), not replacement. AI is less about replacing people and more about letting one person do the work of two or three. For a small business, that means you can serve more customers without hiring more people.
Level 1 (manual use): $0-40/month for ChatGPT and Claude subscriptions. Level 2 (automation): $60-150/month (subscriptions + Zapier/Make). Level 3 (custom solutions): from $500 one-time for chatbot or automation development. ROI is usually positive from the first month: $20 for ChatGPT can save 20-40 hours a month.
Absolutely. AI tools work the same in any country. ChatGPT and Claude understand Ukrainian well. Anthropic Economic Index shows that AI adoption is spreading beyond the top 5 countries, and Ukraine has one of the strongest technical communities in Europe. The only difference is the speed of adoption – which is exactly why early adopters in Ukraine will gain the biggest advantage.
Data instead of hype: your action plan
The Anthropic Economic Index made one thing clear: AI is no longer an experiment. It is a tool that really cuts time, improves work quality, and creates competitive advantage. But only for the people who start using it.
Your plan for the next 30 days:
- Week 1: Sign up for ChatGPT or Claude. Start with one task: copy generation, data analysis, or market research.
- Week 2: Identify 3 processes in your business that repeat every day. Try to automate one of them with AI.
- Week 3: Evaluate the result. How much time did you save? What about quality? What can be improved?
- Week 4: Scale it: add a second tool or automate the next process.
Remember: experienced AI users get 10% better results. Every day you do not practice is a day your competitors build an advantage.
Read also: GEO optimization for AI search ยท How to set up Google Ads from scratch ยท About me ยท Case studies
Comments
Discuss on Telegram