KOSTIVERE — Safe Use of AI: When It Helps and When to Be Cautious

The training takes place as part of the Eesti.ai initiative “The Most AI-Savvy Nation” and supports informed, responsible use of AI in working life.

In this workshop, we will learn to recognise situations where using AI is helpful, requires additional checks, or should be avoided. Through practical examples, we will look at how to give a tool clear instructions, assess the reliability of its responses, and protect our own, clients’ and organisation’s data. AI can help draft emails, summarise texts and find solutions to everyday work tasks. However, a convincingly worded response may contain errors, and a convenient way of working may create data protection or confidentiality risks.

The aim of the training is to give participants the skills to use AI thoughtfully in their everyday work: to choose a suitable task, write a clear prompt, check the result and recognise the main risks. The training is suitable for working-age adults who want to start using AI as a work tool or make their existing practices safer. No prior technical knowledge is required.

By the end of the training, participants will understand how AI and large language models work and what their limitations are; be able to name at least three ways to use AI in their work and distinguish between suitable, restricted and unsuitable use cases; be able to write a prompt that states the goal, necessary background, input, constraints and desired response format; be able to spot fabricated facts, unsupported conclusions and the risk of bias, and know what needs to be checked; understand which data and documents must not be entered into public AI tools; and be able to use a simple safety checklist and understand the responsibilities of individuals and organisations.

**Teaching methods and practical outcome**

Short explanations will alternate with demonstrations, discussion and exercises. We will use sample data that contains no real personal data or confidential information. Participants will create one sample prompt and a safety checklist suited to one of their work tasks. Learning outcomes will be assessed through exercises and group discussion.

**Tools and what you need to participate**

We will use ChatGPT and Google Gemini. The aim of this selection is to offer participants as many freely available AI options and features as possible, so they can apply what they have learned independently after the training. All exercises can be completed using free plans. Prepared sample texts—not real personal data or confidential documents—will be used in the exercises.

Please bring a laptop and charger, and check before the training that you can sign in to both ChatGPT and Google Gemini. If there is a technical issue, exercises can be completed in pairs.

**Trainer**

Natalja Kitam is a specialist with extensive knowledge and experience in psychology, neurolinguistics, psycholinguistics and neuropsychology. She is active in the educational technology sector and is an entrepreneur who creates innovative learning solutions. Natalja also has considerable work experience in the media, including ERR, and in public-sector communications, having collaborated with the Social Insurance Board and the Integration Foundation. She has taken part in large-scale European Union-funded projects and delivered training through the “Settle in Estonia” programme, which supports new immigrants and refugees in Estonia.

The workshop is part of AI Week and the Eesti.ai initiative “The Most AI-Savvy Nation”.