AI adoption in your company

AI adoption, or implementing AI in a company, does not start with buying another tool. It starts with data, rules and people who know how to work with AI safely. We prepare a plan, help you meet the AI Act, guide you through every step and train your people.

In shortAI adoption means bringing AI into a company's everyday work so that it uses the right data, respects access rights and people know how to use it safely. It concerns every company whose people use AI with company data, and it includes obligations under the EU AI Act and the GDPR. Develogi, based in Brno and Zlín, prepares your plan, sets up data, access rights and tools, writes your AI usage policy and trains your employees, across the Czech Republic and Slovakia.

Where AI adoption gets stuck

When did someone last tell you on your first day: “Here is one login. Connect it to AI and you will learn everything about our strategy, our product and what you need to know in your role”? For most people, never.

Companies want to adopt AI, but they lack the basics: data in one place, clear access rights and written-down knowledge. Without them, even the best AI is just an expensive intern nobody briefed.

AI is already in companies. The rules are not.

How to implement AI in your company in 9 steps

Three phases, nine steps and employee training on top. Each step comes with a concrete example of how to do it.

Phase 1: Goals and rules

  1. 01

    Goal and owner

    Management decides what AI should bring the company and names the person who leads the adoption. A project left to IT alone does not get finished.

    How to do itA one-page AI charter: 2-3 goals in hours or money, the program owner, and who approves tools and data.

  2. 02

    Shadow AI inventory

    Find out which AI tools people already use and with what data. Most companies are surprised.

    How to do itAn anonymous survey for all employees. The answers become a register of AI tools: purpose, data, risk and owner.

  3. 03

    Approved tools and rules

    Instead of a ban, give people a safe alternative and clear rules on what may go into AI and what may not.

    How to do itBusiness licenses where the vendor does not train on your data, a one-page AI policy and a form to request a new tool. Default rule: when in doubt, do not paste the data into AI.

Phase 2: Data and access

  1. 04

    Data in one place

    Information must not live only in people's heads, emails and shared drives. Without accessible data, AI has nothing to work with.

    How to do itFor the first use case, list every data source, including spreadsheets and paper files. Whatever AI needs and exists only on paper, move into a system first.

  2. 05

    Data governance

    Every data area has an owner, quality rules and a traceable origin. You know which data is correct and current.

    How to do itSplit data into public, internal, confidential and personal. Give each category an owner and write down what must never go into AI, such as salaries or clients' health data.

  3. 06

    Access rights

    Decide who may see what. AI then shows everyone only what they have access to, and sensitive data stays protected.

    How to do itBefore you connect AI to a drive or SharePoint, review sharing and remove “anyone with the link” access. AI inherits the user's permissions: an accountant cannot see salaries through AI if they cannot see them without it.

Phase 3: Knowledge and operation

  1. 07

    Predefined roles

    A new hire gets a role and with it every access right, tool and piece of knowledge. Assign the role and onboarding is done.

    How to do itThe “salesperson” role includes the CRM, the price list, product materials and an AI assistant with access to those only. A new salesperson is useful from day one.

  2. 08

    Company knowledge base

    Strategy, vision, product and processes written down in one place. The content has an owner who reviews it regularly.

    How to do itStart with one area, such as the product FAQ. Delete duplicates and old versions, assign an owner, then connect the base to AI through an MCP server, read-only at first.

  3. 09

    Pilot and measurement

    Deploy AI first on a few processes where the benefit is visible. Without measurement you cannot tell whether the investment pays off.

    How to do itEach team lists what slows it down the most. Rank the candidates by benefit and effort, pick three and measure the baseline in hours. The pilot is led by someone from the business, not from IT.

10

AI training for companies and employees

Data and rules are not enough if people do not know how to work with AI. We train hands-on, on the employees' own work, so that shadow AI does not take root. Since February 2, 2025, the AI Act also requires companies that use AI to support the AI literacy of their staff. We therefore finish the training with records you can show in an inspection.

  • Shadow AI - why not to use unapproved tools and what happens when company data ends up in a public chatbot.
  • Data protection - what you may put into AI and what you may not, so that data does not leak.
  • Approved tools - which tools the company uses and how to work with them properly.
  • Checking outputs - AI can make things up. People learn to verify a result before they use it.
  • Practice - basic workflows that show people right away how much time AI saves them.

How to do itAn introductory session for everyone on their own non-sensitive documents, then workshops by role. One AI champion in every team with an hour a week to try things and share workflows. Attendance records serve as evidence for the AI Act.

More about AI training for companies

AI Act: what every company that uses AI must have

If your people use AI with company data, the EU AI Act and the GDPR apply to you. That holds even if you have not officially introduced AI and people use it on their own. In practice you need:

  1. Trained people. The company must help everyone who works with AI know how to use it and what its risks are, according to their role. In force since February 2, 2025 (Article 4 of the AI Act). Keep records of the training.
  2. An overview of the AI you use. A list of tools and what they are used for, including shadow AI. The purpose determines the risk: AI in hiring or employee evaluation has stricter rules.
  3. An AI usage policy. Which tools are approved, what must not go into AI and who is responsible for what. Without it, you cannot show that you manage the risks.
  4. Personal data protection. Under the GDPR, the company is responsible for data put into AI, even when an employee did it through a private account. You need tools with a data processing agreement that do not train on your data.
  5. Transparency. When a customer talks to AI or sees content that AI created or altered, they must be able to tell in the cases the AI Act sets out. Recognizing employees' emotions, for example, is prohibited.

We take care of itThe AI tool inventory, the AI usage policy, training with records, approved tools and the setup of data and access rights. You get documentation you can show in an inspection.

Want to start on your own? Download our AI usage policy template in Word.

The full EU AI Act guide for companies

Want help with this?

We make you a plan, guide you through it and train your people.

I want an AI adoption plan

Frequently asked questions

Does the AI Act apply to our company?

Yes, if you use AI in any way with company data, such as ChatGPT or Copilot for texts, emails or analyses. Under the AI Act you are then a deployer of an AI system, and since February 2, 2025 you must support the AI literacy of your staff. Even if you do not officially use AI, employees often use it on their own through private accounts (shadow AI). Under the GDPR, the company is responsible for the company and personal data they put there. An AI inventory, a policy and training are therefore necessary for practically every company.

When does the AI Act apply?

The AI Act applies in stages. Since February 2, 2025, the prohibitions of unacceptable practices and the duty to support employees' AI literacy apply. Since August 2, 2026, most rules apply, including transparency (Article 50), and authorities can supervise and impose penalties. The rules for high-risk systems, such as AI in hiring or employee evaluation, apply from December 2, 2027, and for AI in products from August 2, 2028.

Has the AI Act been postponed? What does the Digital Omnibus change?

Only in part. The Digital Omnibus amendment (Regulation (EU) 2026/1744) has applied since July 27, 2026. It moved the rules for high-risk systems to December 2, 2027 and August 2, 2028, and softened Article 4: a company must take measures to support AI literacy but does not have to guarantee a specific level of knowledge for each person. The prohibitions, AI literacy and transparency have not been postponed.

What does the AI Act require from a company?

Every company that uses AI needs trained people with documented records, an overview of the AI tools in use including shadow AI, an internal AI usage policy, personal data protection under the GDPR, and transparency toward customers where the AI Act requires it. Stricter rules apply to high-risk uses, such as AI in hiring.

What are the fines under the AI Act?

Up to EUR 35 million or 7% of worldwide turnover for prohibited practices, up to EUR 15 million or 3% for breaching other obligations including transparency, and up to EUR 7.5 million or 1% for supplying incorrect information to authorities. Small and medium-sized companies pay the lower of the two amounts. Penalties for missing AI literacy are set by national law.

Who supervises the AI Act in the Czech Republic?

The main supervisory authority will be the Czech Telecommunication Office, which has had an AI supervision unit since September 1, 2026. The Czech National Bank supervises the financial sector and the Office for Personal Data Protection covers biometrics. The government approved the Czech AI Act bill on November 9, 2026, and it now goes to the Chamber of Deputies. For some breaches, the authority is first to issue a warning and a call to remedy them.

Does the AI Act require companies to train employees?

Yes. Article 4 of the AI Act has applied since February 2, 2025 to every company that uses AI. The company must support the AI literacy of its people according to their role. No certificate is needed; you can keep an internal record of the training.

What is AI literacy?

AI literacy is the knowledge and skills that let people use AI in an informed way: they know what AI can do, what its risks are and how to check its output. Under Article 4 of the AI Act, a company must support it according to its people's roles and experience. No certificate is needed; you can keep an internal record of training and other measures.

Who can help us meet the AI Act, train our people and adopt AI?

Develogi, a software company with offices in Brno and Zlín that works for companies across the Czech Republic and Slovakia. We take an inventory of AI tools including shadow AI, prepare an AI usage policy, train employees by role with records for the AI Act, set up approved tools, data and access rights, and guide you through AI adoption from plan to operation.

What is AI adoption?

AI adoption is the gradual introduction of artificial intelligence into a company's everyday work, so that AI uses the right data, respects access rights and people know how to use it safely.

How do we implement AI in our company, and where do we start?

With a goal, an inventory and data. Management decides what AI should bring, you find out which tools people already use, and you write down where the data lives and who may access it. Only then does it make sense to connect AI to the data.

Is it safe to use ChatGPT or Copilot at work?

Yes, if you use a business plan with a data processing agreement where the vendor does not train on your data, such as Microsoft 365 Copilot Chat with a work account, Claude Team or Gemini in Google Workspace. Free and private accounts are no place for company data. A person always checks the output, because AI can make things up.

What is shadow AI?

Shadow AI is the use of AI tools the company has not approved, such as private accounts in public chatbots. It risks leaking company and personal data. According to IBM, a high level of shadow AI adds USD 670,000 to the average cost of a data breach (2025).

Can a company ban ChatGPT for its employees?

It can. The employer decides which tools people use at work. According to lawyers, however, a blanket ban does not work: people keep using AI, just in secret through private accounts. It is safer to approve a company tool and give people clear rules.

What should an AI usage policy contain?

Which AI tools are approved, what data must not go into AI, who is responsible for what, how outputs are checked, how incidents are reported and how people are trained. You will find a 14-section policy template to download in Word on our blog.

What can go wrong with AI in a company?

  • Data leak. Samsung engineers pasted confidential source code into ChatGPT, and the company then banned generative AI on company devices (2023). TechCrunch
  • Made-up filings. The Czech Constitutional Court fined a lawyer CZK 25,000 for a complaint citing case law that AI had invented (2025). Seznam Zprávy
  • Hallucinations in a contract. Deloitte refunded the Australian government more than AUD 97,000 for a report with AI-fabricated citations (2025). CFO Dive
  • A chatbot promised too much. Air Canada had to honor a discount its chatbot wrongly promised a customer. The tribunal ruled that the company is responsible for its chatbot (2024). The Guardian
  • An AI agent without limits. An AI agent deleted SaaStr's production database even though it had been told not to make changes (2025). Fortune
Rules, approved tools, limited access rights and output checks would have prevented every one of these cases.

How often do AI incidents happen in companies?

Often. In 2025, 13% of organizations reported breaches of their AI models or applications, and 97% of them lacked AI access controls (IBM). Generative AI is involved in 32% of data security incidents (Microsoft, 2026). 34.8% of the company data people put into AI tools is sensitive (Cyberhaven, 2025). Damien Charlotin's databaselists more than 2,200 court decisions that dealt with content made up by AI (November 2026).

What is an MCP server?

MCP (Model Context Protocol) is an open standard that connects an AI assistant to company data and tools, such as accounting, a CRM or a company wiki. When the server respects access rights, the assistant shows everyone only what they have access to.

Why do AI pilots often fail to reach production?

They lack a connection to real data and processes. According to MIT NANDA, only 5% of companies got specialized AI tools from pilot into production. What helps is an owner from the business, a baseline metric and a plan to connect to production systems from the start.

How long does AI adoption take?

It depends on the state of your data and rules. After an initial analysis, we give you a plan with individual steps and dates.

Cookie settings

Necessary cookies keep the website working. With your consent we also use analytics cookies (Google Analytics) that help us improve the website.

Necessary

Always active

Website language, spam protection of the contact form (Google reCAPTCHA) and storing your cookie choice.

Website visitor statistics using Google Analytics.