Short answer: your staff are probably using AI already, and many of them were never shown how to use it well or safely. The fix is a short, practical programme built on each team's real work: a few tasks per team, clear rules on what data stays out, and a habit of checking the output. A free government course can cover the basics first.
People are using it. Few were shown how.
In an online survey of 1,704 professionals run by Genius HRTech and Digipoll in November 2025, 67% said they already use AI tools to simplify or automate daily work. Only 37% said they had received proper training, and 61% said their organisation had not given adequate guidance on how to use AI well.
Simplilearn's 2026 Professional Sentiment Survey, which covered professionals across countries and industries, found the same pattern: 85% use AI regularly at work, while 71% say their employer is not preparing them adequately.
That leaves staff to figure it out alone, often on personal accounts. Back in 2024, Microsoft and LinkedIn's Work Trend Index found that 78% of AI users were bringing their own AI tools to work, and only 39% of people using AI at work had received training from their company.
For an owner this is a quality problem and a data problem. Two people in the same team get very different results from the same tool, and nobody has told either of them what customer information they should keep out of it.
Use the free course for the basics
In November 2025 the IT ministry launched YUVA AI for ALL, a free, 4.5-hour self-paced course in six modules under the IndiaAI Mission. It is on FutureSkills Prime, iGOT Karmayogi and other platforms, and anyone who completes it gets a Government of India certificate. The modules cover what AI is, how it is changing work, using AI tools safely and responsibly, and Indian examples.
It is a sensible week-one assignment for everyone, including people who say they don't need it. It gives the whole team the same vocabulary. What it can't do is teach your sales team how to answer your enquiries or your accounts team how to reconcile your vendor statements. That part you have to build.
A training plan for a 20 to 200 person company
1. Pick the tasks before the tools. Ask each team lead for two or three tasks that eat hours every week and follow a pattern: first replies to enquiries, meeting notes, a weekly report, drafting a quote, reading supplier invoices. Training on those tasks gives people something they can use the next morning.
2. Write the rules down first. Before anyone practises, decide which tools are approved, which accounts to use (company, not personal), and what never goes in: customer phone numbers, salary data, client contracts, passwords. Put it on one page. If you don't have that page yet, we wrote about what to put in a small company's AI use policy.
3. Teach checking as its own skill. In the Genius HRTech survey, 49% said they trust AI-generated insights without manual cross-verification, while 36% said they don't trust them. For each task, agree what must be verified (numbers, names, dates, anything sent to a customer) and how. A quick rule helps: if it leaves the building or changes money, a person reads it first.
4. Run short sessions on real work. Thirty to forty-five minutes per team, using that team's actual documents with sensitive details removed. Show one task, let everyone try it, and compare the outputs side by side. People learn more from seeing why a colleague's version came out better than from any slide.
5. Name one or two champions per team. There is usually someone already ahead. Give them time to help others and to collect the prompts and steps that work, saved in one shared place so they don't live in personal chat histories.
6. Measure one number per task. Time to first reply, hours spent on the weekly report, errors caught in review. Check it before training and four weeks after. If the number didn't move, change the task or the method before training the next team.
What to skip
Long generic workshops with no follow-up. Certificates as the goal; they show someone sat through a course, not that their work got faster. Buying a new AI tool for every team before people are using the one you already pay for. And training only the junior staff: when managers don't use the tools themselves, they can't judge the work that comes back.
Where to start this month
Send everyone the free course link this week. Pick one team and two tasks, write the one-page rules, and run the first session on real work. Measure, adjust, then move to the next team. A month of that is a better start than leaving people to work it out alone.
Arthat AI
Working on something similar at your company?
Arthat AI builds these automations for Indian businesses every week. If you want to talk through whether this fits your situation, we'll give you a straight answer — even if the answer is "don't build it."
