Digitalisation in the forest industry is advancing rapidly, and the volume of data is growing just as fast. A new artificial intelligence (AI) assistant provides easy access to verified forestry expert knowledge, using language that is natural and understandable to everyone.
The Finnish Forest Products Engineers’ Association has created ForestBioFacts – a global digital learning platform for the forest industry – where a new AI-based solution is now available. ForestBioFacts is the leading digital learning platform in forest bioeconomy, containing 1,700 articles across 16 topics. The platform is used by more than 50 organisations worldwide. The new ChatGPT-style learning tool for the forest industry introduces a conversation-based interface to the platform, allowing users to obtain information on forest sector topics directly through AI, making the knowledge base more interactive.
“Finding information is no longer based on search terms alone, but on conversation. Users can ask questions, clarify details and deepen their knowledge much as they would with a person – an expert,” notes Vera Hallikainen, development projects manager at the Finnish Forest Products Engineers’ Association.
Reliability of information matters most
The key premise of the solution is reliability. The content of ForestBioFacts is built on a broad network of experts. Around 200 experts from companies, research institutes and universities in Finland and elsewhere in the world have contributed to its development. The knowledge base includes the Papermaking Science and Technology series, written by industry specialists.
“The goal is to provide information that is both quickly accessible and reliable, and understandable in context,” stresses Vera Hallikainen. According to her, artificial intelligence also improves how quickly information can be identified.
“Users can ask questions in natural language, ranging from individual terms to broad topics, while at the same time building learning paths tailored to their own needs. Finding information is no longer based solely on search terms, but on conversation. Users can ask questions, clarify details and deepen their knowledge just as they would with a human expert,” says V. Hallikainen.
The forest sector’s ChatGPT
Unlike open-web systems, the recently introduced AI agent draws its information only from the ForestBioFacts learning environment. “In the ForestBioFacts model, answers are based on curated content, and the user always sees which sources the information provided has come from,” explains V. Hallikainen. “This ensures the reliability of the answers and the traceability of the information. A registered user can ask questions about forest industry technologies, products and processes and receive answers based on expert-prepared information, which also includes links to sources.”
In recent years, the reliability of generative artificial intelligence systems has been widely discussed, and it is a significant issue. For example, the journal Nature has reported in several articles that large language models can produce statements that sound credible but are incorrect, without citing sources. MIT Technology Review describes this phenomenon as “hallucinations”, in which AI fills gaps in information with its own guesses. MIT Technology Review has called it the main obstacle to wider adoption of chatbots.
An incident in Finland in spring 2026 illustrated the consequences of such hallucinations. Two major news agencies published an erroneous report about drones said to threaten national security. Both agencies stated that the error stemmed from incorrect information generated by an AI tool. In an opinion piece for Kaleva, Jorma Laitinen, a doctor of social sciences and lecturer in social policy, argued that the incident could reflect a broader systemic problem rather than isolated mistakes. “AI programs are already so advanced that, despite increased oversight and fixes to the tools, we can never be fully certain that they will act solely in line with our objectives in every situation,” Laitinen wrote.
In this context, the AI assistant developed by ForestBioFacts for the forest sector represents a different approach. It is built on a closed, expert-reviewed and curated knowledge base, and generates answers only from selected content, with a clear source reference attached to every answer. This increases transparency and addresses concerns about AI reliability.
Supporting skills development and access to knowledge
AI-supported learning environments can significantly improve skills development, access to information and decision-making. At the same time, the reliability of generative AI has sparked debate, particularly regarding solutions based on open online sources.
V. Hallikainen explains that the ForestBioFacts AI assistant addresses this problem by basing its answers on expert-assessed content and citing sources for every answer. Solutions of this kind, built on a limited, quality-assured database, remain rare in Europe. Most generative AI systems draw on large, publicly available datasets that may contain inaccurate or unverified information, which increases the risk of factual errors.
In practice, AI makes finding information easier. During a conversation, a huge amount of material becomes accessible without traditional searching, and users can refine their questions as needed. This helps students, researchers and professionals alike.
According to V. Hallikainen, AI helps both with finding detailed technical information and with understanding the broader context. It can also support the creation of personalised learning paths. This solution reflects a wider trend in which AI is increasingly paired with a limited and trustworthy knowledge base. This approach is especially important in fields where incorrect information can have serious consequences.
“Artificial intelligence is already being used in the forest industry to streamline processes and optimise production, but its role is also growing in data management, research and innovation. The ForestBioFacts AI agent reflects a trend in which AI supports expert work and decision-making,” explains Vera Hallikainen.
Looking ahead, AI is expected to particularly strengthen continuous learning. Learning will become more personalised, information will be tailored to users’ needs, while expertise remains the core element. Development of the forest-industry-focused AI agent is currently ongoing.
“The next step will be to expand the knowledge base by including historical publications and teaching materials. This autumn, around 113,000 pages of new content will be added to the system, searchable with the help of AI,” notes Vera Hallikainen.