AI-contracts training for lawyers

After two sessions, your legal team understands the background of AI regulation and the technology itself, and can add more value to AI projects than before.

In the training, you will learn:

Market practice for procurement and sales contract terms for AI systems
The effect of AI technology on contracts
Consulting and subcontracting contract terms that account for the use of AI
The effect of AI maturity on contract clauses

AI contracts demand combined expertise: you must understand law, technology and business strategy at the same time.

Why this training?

A year ago, I decided to write a book about AI contracts. Before writing it, I thought I understood AI law. I had read the regulations and helped clients draft contracts. Even so, several contracting situations were difficult in a way I could not put into words precisely.

I understood the problem only once I started developing AI services myself — when I began to see what service providers actually sell and how services and software are used. I also saw how the promises of organizations’ pilots went unfulfilled despite high hopes. That rarely came down to the technology failing. Instead, organizations were not ready enough to adopt new ways of working. This showed in culture, data quality, staff competence, and processes. It also showed in the contracts and guidelines that companies used. Contracts have to change to manage the new kinds of errors that AI systems make, and guidelines have to carry the agreed terms into practice. Designing contracts for the AI era requires the lawyer to understand which AI options exist and to shape the contracts to match those options.

After talking with dozens of organizations, it became clear that lawyers’ technical and legal competence does not determine success on its own. To succeed, lawyers must work with management teams to develop strategy, guidelines and a roadmap — so that AI contracts advance staff competence, data capability and the strategic vision of the whole organization.

The training promise

I wrote the book Tekoälysopimukset (Alma Insights 2026 — together with Hanne Hirvonen), and I train legal teams and management teams on the subject. My time is limited, and I run the training properly. For that reason, I will hold only six trainings in 2026. Because buying training takes up valuable time and costs money, I want to carry the risk of its success. That is why I decided to give three promises.

Promise 1 — Every participant learns something concrete

Every participant learns at least one thing they can use in their work within the next month — a question to put to a supplier, a clause to add to a contract, a new risk they can flag and manage, or an old practice they can drop.

According to a recent study by Suomen Ekonomit (5/2026), not every workplace yet has adequate guidelines or training on the correct use of AI in work tasks — and competence gaps can have critical consequences. At the start of the training, I take time to make clear what AI models technically are and what they are not. Technical understanding is the foundation on which all other understanding of AI law is built.

The goal of the training is for the legal team to bring safety and added value to AI projects — not needless braking.

If the training does not meet this promise, I refund its price.

Promise 2 — Contracts faster, advantages secured

After the training, your legal team speeds up the negotiation of AI supplier contracts by a full week — and at the same time secures the organization’s interests better than before. When the legal team knows which questions belong in a contract negotiation and in what order, the negotiation no longer stalls on internal deliberation.

After the training, lawyers understand what is worth negotiating and which matters contracts cannot resolve. Cooperation between lawyers and technical experts becomes easier, and legal work aligns with the progress of AI projects in the right way.

Promise 3 — AI projects succeed

According to research, the success rate of AI projects is only about 5%. My aim is to help my clients quadruple this chance of success.

The conditions for success are often set already at the contract level. According to the same MIT study, buying AI services from specialized suppliers succeeds in about 67% of cases, while solutions companies build themselves succeed only about a third of the time. The right choice of supplier and a carefully negotiated contract therefore decide a significant part of a project’s outcome — often before a single line of code has been written.

Well-drafted contracts set the right expectations, protect data, establish monitoring mechanisms, and provide a framework within which projects produce value. In poorly drafted contracts, expectations, metrics and responsibilities remain vague — and then the project struggles to succeed.

The big promise — the whole team on the same level

This training was not a walkthrough of the AI Act, which you can buy elsewhere too. We would not have gained this kind of competence.

Training feedback

In large companies there is often a considerable gap within the legal team in how well AI regulation is known. One or two lawyers have read the AI regulations; the rest have not yet. At the same time, according to surveys, every team member understands that AI has already changed work.

The goal of the training is for the whole legal team to understand why AI is regulated in a particular way and to apply that understanding to their own contracts. The training builds more than knowledge of the statutes — it builds the ability to justify contract requirements to the client and the supplier in a sensible way.

Download the first chapter of the book (PDF)

What does the training consist of?

The training comprises two half-days, each three hours long. You can order both or just one — it depends on your situation regarding AI.

Part 1 — Buying AI services

If your organisation procures or plans to procure AI services — chatbots, translation tools, learning systems, automatic document processing, risk-assessment models or anything else — this part is meant for you.

We cover:

The AI Act’s obligations for buyers and users concretely: what the AI literacy obligation (Article 4) requires of you, what the obligations of a high-risk system’s deployer mean, and how you identify whether a system falls into the high-risk class.
Checklists for supplier due diligence: what you ask at the buying stage, what you check in the contract, how you document the reasons for the procurement decision.
Reconciling data protection and AI in procurement — when automated decisions require human review, when a new register arises for which an impact assessment should be made, and how responsibility is divided with the supplier.
Copyright, intellectual property rights and rights to use data. AI-related copyright questions differ significantly from traditional software licensing, which makes them difficult to address without specific technical expertise — including what terms your data may be part of the training data for the supplier’s model, who owns the end product the AI produces, and how confidential information is protected in systems that use AI. In the training, I go through how these questions are handled concretely in contracts.

Part 2 — Developing and offering AI services

If your organization develops AI solutions for its own use, sells them, or offers them to its customers — either as standalone products or as part of another service — this part is meant for you.

We cover:

The AI Act’s obligations for providers: product notification, documentation, post-market monitoring, conformity assessment.
Regulatory considerations in product development: privacy by design, data governance, model selection, and the impact of model architecture on legal risks.
Liability questions at the customer interface and the insurance angle: who is liable if your AI service causes harm to a customer.
End-customer contracts and service-level questions — how your contracts hold up when an AI model changes, updates, or goes wrong.

Who is the training built for?

This training is designed for the legal teams of large companies, public-sector organizations and associations, where AI has become a visible topic but competence is not yet evenly distributed.

The typical situation the training is built for: the team has one or two lawyers who have read the AI Act to some extent. For the others, the AI Act is still something that has been talked about but opens up a little awkwardly. At the same time, everyone agrees that AI is changing their work. From this starting point, the training lifts the whole team onto the same level.

The training also suits a management team, a data protection officer, and a procurement function — especially if representatives of these roles take part in the same training as the legal team. A shared language is built best together.

Who is the training not for?

If your legal team already specializes in AI full-time — that is, all of its members work with AI law every day — this training does not meet your needs. The training does not benefit a team in the legal unit of a company that develops AI as its main business.

The training is also not suited to these situations:

Are you looking for an academic presentation of the AI Act or of AI regulation? Universities and commercial training providers offer those. This training focuses on practical action — on what you should do to your own contracts and processes tomorrow.

Are you looking for a general, short overview of the legal questions around AI? This is a deeper package in which the discussion centers on your own contracting situations. During the training, I also do not present or sell other services — the training is already a paid whole.

Do you need an online course you can complete at your own pace? All my trainings are real-time and interactive. The training is ordered for a team because discussion and learning from others are essential parts of it — a one-person training would not produce the same result.

The training does not work well as a mere compliance checkbox. It will change thinking and challenge ways of working.

Can contract work and training be combined?

Yes. The best benefit comes when your contracts are revised at the same time to account for the new risks AI causes, the background of the regulation and the new types of clauses — and when these updated terms are gone through with the team in the training.

Then the training does not revolve around abstract model contracts, but addresses the contract templates your team actually uses. Examples of the combination: developing AI-related contract clauses for framework agreements, assembling a clause bank for different situations (your own procurement, subcontractors, product development, customer contracts), tailoring model contracts to your industry, building AI guidelines for staff — and then a training in which the legal team internalises the new terms and learns to apply them.

The scope and price of the contract work in such development projects are agreed on a case-by-case basis.

What does the training cost?

The two-day (3+3h) training package costs 8,000 € + VAT. The price covers both half-days — six hours of training in total. It includes preparation, materials, delivery and the time I spend in advance mapping your situation.

A single half-day costs 5,000 € + VAT. With the standard package, you save 2,000 €, because the preparation work is the same size whether one or two half-days are held. Order the two-training whole if you can.

The price is justified because a well-drafted AI supplier contract delivers measurable benefits throughout the contract period. In AI procurement, mistakes and fixing them can prove considerably more expensive than careful preparation.

Schedule and capacity

In 2026 I will hold only six tailored trainings. Tailoring takes so much time that I cannot expand the offering without compromising the quality promise.

The first free slot is in September 2026. If the calendar is already full for the rest of the year, we propose a time for the following year, or we consider together whether the training can be staged to fit both schedules.

What happens after the training?

Your legal team and management team identify the AI Act’s key obligations in your own organisation, and can tell apart the roles of deployer, user and provider.

The legal team knows the right questions to ask suppliers at the procurement stage of an AI service — the questions that separate careful due diligence from last-minute decisions. The team understands the technological basics of AI models with enough precision to recognise how technology choices affect legal liability questions: whether to build your own model or use a ready-made one, whether to fine-tune a licensed model or use it as is, and what the different architecture choices mean from the perspective of contractual risk.

You get tools to use as checklists for contract clauses and as a basis for your organisation’s internal AI guidelines. You also gain a sense of the strategic use cases for AI and tell them apart from cases where AI mainly produces risk or cost.

The materials stay with you afterwards: the slide set, the contract models and summaries of the key statutes. You can use these in your team at no extra charge after the training as well.

What have participants said?

The numbers from participant feedback:

Recommendation score 8.2 / 10
Relevance of AI to one’s own work 4.7 / 5
Trainer’s expertise 4.2 / 5
Need for further training 4.2 / 5

A quote from the feedback analysis:

“The trainer was very knowledgeable and explained AI-related phenomena especially clearly.”

Our experts are ready to answer your questions

Leave a call-back request

Book a meeting

Request a quote

Previous training clients

Reson trains AI law regularly for lawyers, engineers, and organizations’ teams. Previous training clients:

Tekniikan akateemiset (TEK) — AI law training for the membership
Moomin Characters — AI and copyright training
City of Espoo — AI law training
South-Eastern Finland University of Applied Sciences (XAMK) — AI Act guidance and training for the university’s staff (read the case study)
Alma Insights — chair and four-time speaker at the AI Law event
A Finnish international listed industrial company — AI Contracts training for the legal team, 2026

Reson’s lawyers have also taken part in preparing AI legislation and advised both organizations developing AI services and those adopting them, in dozens of assignments over the past five years.

Who runs the training?

The training is run by Herkko Hietanen, founding partner of Reson Oy:

Master of Laws (OTM) — thesis on open-source licensing
Doctor of Science in Economics (PhD) — dissertation on copyright licensing and business models
Master of Science in Technology (DI), software engineering and business — on the problems of integrating public information systems

Experience:

Licensed trial counsel, bar examination completed
EUIPO IP Scan service provider
ERMI expert (copyright disputes, appointed by the ministry)
24 years of work experience in IT and technology law
Over 100 copyright proceedings and over 200 assignments in the past five years
Research background: HIIT (Helsinki Institute for Information Technology, 2001–2012), visiting researcher at Harvard’s Berkman Center, MIT’s Communications Futures program and UC Berkeley
Software development background — has built applications in client projects and won programming competitions (including Slush, Apps for Finland)

I have written the book Tekoälysopimukset (Alma Insights 2026), which serves as the content basis for the training.

What can you download for free right now?

Before you book a meeting or request a quote, you can read the first chapter of the book Tekoälysopimukset. You get it without an email, without registration, without extra marketing messages in your inbox.

Download chapter 1 of the book (PDF)

Chapter 1 introduces the subject: what an AI contract really is, how it differs from a traditional IT contract, and on what principles it makes sense to approach AI contracts.

If you want to discuss the training after reading the chapter, we can book a meeting, or you can request a quote. If, on the other hand, you do not recognize your own situation in the table of contents or the introduction, the training probably does not meet your needs.

FAQ

Can the training be combined with contract work?

Yes. If you want the training to use your own contracts as examples, we can do the contract work first and the training after. This is the recommended approach, especially if your organization is currently building a playbook or framework agreements for AI contracts.

Can you get a refund if you are disappointed?

Yes. If the training does not meet Promise 1 — that every participant learns at least one thing they can use in their work within the next month — We refund the price.

Can the training be held entirely remotely?

Yes. I also hold trainings over Teams or a similar remote connection. The biggest difference from on-site training is the rhythm of discussion: remotely, the discussion is more compact and requires more facilitation from the host to keep people engaged. I recommend on-site delivery when it is possible, but remote delivery works well too.

Can I order just one of the half-days?

Yes, a single half-day costs 5,000 € + VAT. The buyer-side training (Part 1) works well on its own if your organization procures AI but does not offer it onward. The provider-side training (Part 2) works well on its own if you operate as a developer or reseller of a technology product.

How many participants are taken?

An optimal group size is 10–20 participants. Below this, it still works, and the discussion is tight; above this, the group becomes so large that interaction weakens. If your legal team is larger, we recommend splitting the group into two cohorts.

Do you get a certificate from the training?

You receive a participation confirmation that can be attached to the documentation for the AI Act’s Article 4 AI literacy obligation.

What do the materials contain?

The slide set (PDF), the script, the contract models, and statute summaries. The materials are also available to the client at no extra charge after the training.

Are contract templates part of the training?

Yes, contract templates are gone through in the training. They are examples — tailoring them to your needs is possible either as a follow-on assignment or as a combination of contract work and training (see the section “Can contract work and training be combined?” above).

How soon can you start?

Typically about 4–6 weeks from order. This includes an initial mapping, in which we determine the participants’ starting level and your key use cases, and the tailoring of the materials to your industry. The first free slot in 2026 is in September.