How is Mondi using AI?

In its annual reporting, so far mainly as a subject for its board, with production named as the focus. Outside that reporting, it already uses AI in at least two places: a university-built tool for choosing paper grades, and HR processes since 2019. The 2025 integrated report records that the audit committee "had an in-depth teach-in on artificial intelligence (AI)". There was "a focus on how AI can be used to optimise Mondi's production processes". The half-year results for 2026 don't mention AI. A November 2025 university release says one Mondi paper business "is already using the software" from an AI research project "to select paper grades for special applications".

What the usual answer says

The top results are trade coverage of Mondi's 2026 e-commerce report, which is about how shoppers use AI, not how Mondi does. A trade item on the paper permeability tool comes next.

What the reports say

The 2024 integrated report mentions AI once. The audit committee's half-yearly IT updates covered cyber defences "and Mondi's approach to the use of AI".

The 2025 integrated report gives AI its own place in the committee's work, under "IT, cyber security and AI". The teach-in covered "how prepared the organisation is for AI and where the greatest opportunities are". It notes the insight "will provide useful context for the committee, and the Board, during its future discussions and decision-making, with AI becoming an increasingly important consideration". The Board "regularly discusses the opportunities that further digitalisation, and AI in particular, can offer". In operations, the report names one data project, "automated digital tracking and verification of certified recycled content, including pilot projects with customers", without calling it AI.

The half-year results announcement for 2026, published in July, has no mention of AI, machine learning or digitalisation.

What sits outside the reports

The university's release, carried by TechXplore on 20 November 2025, describes an AI-based system that predicts how permeable different papers are to volatile substances. It combines lab data with physics-informed neural networks. One of Mondi's paper businesses took part in the research lab. The release states that it "is already using the software to select paper grades for special applications". That is AI in product selection, in use, from a research partnership.

Earlier still, a 2019 Mondi article describes its HR system. "All the information is stored in such a way that AI can filter out applications meeting the defined requirements." It also prompts managers to hold exit meetings. It is seven years old, but it shows AI in an operational function well before the board teach-in.

Mondi's 2026 e-commerce trend report, announced on 27 May 2026, is about customers' shopping behaviour and says nothing about AI in Mondi's own operations.

The scale behind the board question

The 2025 report describes a group running 14 mills and "over 100 production sites", with 24,000 people. In our view, that is why the board's question about production is the right one and also a hard one. A model that optimises a paper machine needs machine, quality and order data that match across sites, and the report doesn't say how far they do. The recycled content tracking pilot is the one place the report shows data being linked from input to output with customers involved.

What it means

An integrated report is a governance document. It records what the board discusses, and AI tools in a lab or an HR system are not material to it. So the reports' picture, AI as a production question for the board, and the partner's picture, AI already choosing paper grades, are both true. For other operators, the board question has a practical answer. Which production data is complete, consistent and trusted enough to optimise on, and where do machine, quality and order records already connect? The permeability tool shows the pattern: a narrow question, with lab data behind it, answered by a model. It replaces a slow test with a prediction that can be checked against the test, which is why it could move into use before any production-wide programme.

When it doesn't apply

Companies already reporting AI in production have moved past the board-question stage. For any company, reports and partner releases each show only part of the picture.

Quarri for pulp, paper and packaging is built around how a fibre operation runs, from furnish to converted order.

Quarri is an AI-native data platform for the timber supply chain. It connects buying, production, sales and inventory for forest management, sawmill, wood products and pulp, paper and packaging operators.