, Nils-Erik Jansson
The general counsel's AI report to the board: two pages
The report is two pages in seven sections and is submitted once a year. Under Swedish company law, the board is responsible for the company's organisation and for seeing that the business and its risks are controlled. AI is one of those risks. When the question goes to the general counsel, who owns neither the technology nor the budget, the report needs to show where the line is. Here is what the report should contain, what the general counsel can stand behind in it, and what belongs with IT and the CFO.
Updated by Nils-Erik Jansson
Why the board asks
The board is responsible for the company's organisation and the management of its affairs, and must ensure that the organisation is designed so that the company's financial position is adequately controlled [1]. The Swedish Corporate Governance Code, which applies to listed companies, goes further. The board must ensure that there are appropriate systems for monitoring and controlling the business and its risks, and satisfactory control of compliance with laws, rules and internal guidelines. The board is also responsible for the company having good internal control [2].
The use of AI is now one of those risks. Thomson Reuters reports that the share of general counsel naming technology as a strategic priority doubled going into 2026, and that nearly half of legal departments have a generative AI tool for the whole company [4]. In July 2026, Axiom found that two thirds of legal departments run general purpose AI tools in their default configuration, and that 83 per cent cannot show whether last year's AI spending produced anything [3]. The report should confirm or contradict that picture for your own company.
The question belongs partly with the general counsel, because it concerns rules, contracts and responsibility, and because the terms are a legal document. But the general counsel owns neither the technology nor the budget. The report should be honest about that line.
Two pages, seven sections
Section one: tools in use. A table with one row per tool, as below. It is Annex 1 of the AI policy, see the guide on an AI policy for the legal department.
| Tool and vendor | Plan and agreement | Users | Owner |
|---|---|---|---|
| name and vendor | plan and under which agreement | number and which parts of the company | named person |
If no such table exists, the first message to the board is that it is missing, and that is a result in itself. Personal accounts used for work are listed as one row: known extent, action in progress, see shadow AI.
Section two: governance. Is there a policy, when was it adopted, who owns it, when is it reviewed? Is there a list of what may be entered and a list of matters kept out? Have staff been trained, when, and how is it documented? Three lines per question. The board wants to know the structure exists. It does not want to read it.
Section three: data protection and contracts. For each tool in the table: have the five questions been answered (data processing agreement, transfer, retention, requests from public authorities, matters kept out), yes or no, with dates. Remaining risks in one sentence each. See the checklist. This is the section the general counsel answers for in full.
Section four: incidents. The number of events during the year, how many were assessed as personal data breaches, how many were notified and what was changed. Zero incidents is an answer that requires an explanation of how you would have found out. See the guide on AI incidents.
Section five: the regulatory position. How the EU AI Act applies to the company: which requirements apply now (AI literacy) and which are coming (transparency, and possibly high-risk if tools are used for decisions about employees), with dates and reservations. More in the guide on the EU AI Act for in-house counsel. For law firms: the Swedish Bar Association's guidance. For regulated sectors: the supervisory authority's position. Half a page at most.
Section six: cost and benefit. Licence and implementation costs in money, from the CFO. The benefit in the measures the department can actually produce, see the guide on five metrics, and honesty where the measures do not yet exist. Axiom's finding that most cannot show whether the spending produced anything [3] is a good reason to set the baseline before the next purchase, and to tell the board so.
Section seven: decisions needed. Three at most. For example: approve the policy, decide on an Enterprise agreement for the whole company, set aside budget for training, or decide that tools may not be used for decisions about employees until governance is in place. The board should leave the meeting with something to decide. Otherwise the report was only information.
What the general counsel answers for
Sections two, three, four and five in full. Section one together with IT, which knows what is actually installed. Section six together with the CFO, who has the figures. Section seven together with the CEO. Say so in the report: who stands behind which part. A report in which the general counsel makes claims about technology and money that they cannot stand behind undermines what they can stand behind.
Mistakes that make the report worthless
The report is about AI in general, with trends and forecasts, when it should be about the company's tools and the company's risks. It lacks the table in section one, so the board does not know what it is deciding on. Remaining risks are hidden in phrases such as "managed on an ongoing basis". It does not end in decisions. It runs to twenty pages when the board reads two.
How HOVR Legal works with this
Sections two to five of the report come out of the data protection package and the AI operating model, because they build on the same documents: the policy, the lists, the assessments and the incident log. HOVR Legal writes the first report together with the general counsel as part of the engagement, in your template, so that next year's report is an update and not a project.
Frequently asked questions
- Is the general counsel responsible for the company's use of AI?
- No. The board is, and under the board the CEO. The Swedish Companies Act places responsibility for the company's organisation and for adequate control of the business on the board [1]. The general counsel reports on the part that concerns rules, contracts, data protection and risk, and makes sure the question is asked. IT reports on the technology and the CFO on the money. The report works best when the three each write their part and the general counsel puts it together.
- Does the Corporate Governance Code apply to us?
- The Swedish Corporate Governance Code applies to companies whose shares are traded on a regulated market in Sweden [2]. Its requirement that the board ensures there are systems for monitoring and controlling the risks of the business and compliance with laws and internal guidelines is still a good description of what every board is expected to do, and that is what the report responds to. An unlisted company can use the same outline.
- How often should the report be submitted?
- Once a year as a standing item, and in addition when something has happened: a breach notified to IMY, the Swedish Authority for Privacy Protection, a new tool for the whole company, or a regulatory change that requires a decision. More often than that becomes noise. Less often means the board finds out from the press.
Statements about Anthropic's terms and products were read on Anthropic's pages on 10 November 2026 and are checked at the start of every engagement.
Sources
- [1]Aktiebolagslag (2005:551), 8 kap. 4 §, Regeringskansliets rättsdatabaser, in Swedish; the Swedish Companies Act, Chapter 8, Section 4
- [2]Svensk kod för bolagsstyrning, gällande från den 1 januari 2024, Kollegiet för svensk bolagsstyrning, in Swedish; the Swedish Corporate Governance Code in force from 1 January 2024, rules 3.1 and 8.1
- [3]Legal AI is everywhere, but only 7% of legal teams have made it work, Axiom, press release 9 July 2026
- [4]State of the Corporate Law Department Report 2026, Thomson Reuters Institute, 24 March 2026
- [5]CLOC 2026 State of the Industry Report, press release 2 March 2026
Insights on the topic
Back to AI for legal
More under AI for legal
- Data protection package for AI tools in legal work
Fixed-scope engagement: transfer impact assessment, impact assessment where needed, draft addendum, policy and training. For the AI tool you use or consider.
- AI operating model for the legal department
Fixed-scope engagement for legal departments. You get an operating model for how AI is used, governed and followed up, whatever the tool.
- AI literacy training for lawyers: two hours
Two hours of training for your legal department or firm on your own tools and documents, documented as a measure under Article 4 of the AI Act.
- Data protection checklist for generative AI in legal work
A checklist for law firms and legal departments: five questions to answer before the first client data goes into an AI tool, and what counts as a pass.
- AI policy for the legal department: what it should contain
Outline of an AI policy for a legal department or law firm: permitted tools, what may be entered, review, responsibility, excluded matters and training.
- AI policy template for the legal department
An editable AI policy template for a legal department or law firm in nine sections, with comments on what to adapt and why.
- The EU AI Act for in-house counsel: what applies to users
What the EU AI Act requires of a legal department that uses generative AI: AI literacy since 2025, transparency since 2026 and when the high-risk rules apply.
- Transfer impact assessment for AI tools, step by step
The six steps of a transfer impact assessment for a US AI tool: map, choose a transfer tool, assess the country, supplement, document, review.
- DPIA for generative AI: when it is required and how
When a legal department or law firm must carry out a DPIA under Article 35 GDPR for an AI tool, using the Swedish regulator IMY's nine criteria.
- Personal data in prompts: legal basis and information
When a lawyer writes about an employee, a counterparty or a customer in an AI tool, personal data is processed.
- AI incident response for legal teams
What a legal department does within 72 hours when a member of staff has entered data into an AI tool that should not have gone in.
- Anonymise before prompting: what works
How a lawyer removes personal data before text goes into an AI tool, and the difference between anonymisation and pseudonymisation.
- Swedish MBL and AI tools: must you negotiate first?
Swedish law: when an AI tool triggers the duty to negotiate under section 11 MBL, when informing under section 19 is enough, and with no collective agreement.
- Shadow AI in the legal department: one week of work
Staff use ChatGPT and Claude on private accounts for work. The legal department's week: map it, stop the worst, offer an approved tool, write the rule.
- Legal privilege and AI: the Swedish Bar guidance
What the Swedish Bar Association's guidance on generative AI requires for confidentiality: tools that do not train on client data, and terms read in full.
- Contracts with AI vendors: asking beyond standard terms
What a legal department or law firm asks of an AI vendor beyond the standard terms: authority requests, sub-processors, deletion, anonymisation, exit.
- On-premise or cloud AI for legal work
When a legal department or law firm should run AI locally and when the cloud is right: what local operation solves, what it costs, what it leaves open.
- Reviewing an AI answer in ten minutes: five steps
How a lawyer reviews an answer from Claude, Legora or ChatGPT before using it: sources, case law and statutes, figures, what is missing, and tone.
- AI literacy training for the legal department
What a two-hour training for lawyers using generative AI covers: what the tool does and does not do, your rules, practice on your own documents.
- AI guidance for lawyers in the EU and the UK
What the CCBE, the SRA, the Law Society, the Bar Council and the Swedish Bar Association require of lawyers using generative AI, and where they differ.