Quebec’s national library is embarking on an ambitious initiative to create a comprehensive database that encompasses cultural and governmental content, aimed at enhancing the performance of artificial intelligence systems in understanding the province’s unique society, culture, and Indigenous languages. The Bibliothèque et Archives nationales du Québec (BAnQ) has launched the experimental phase of this project following a feasibility study completed earlier this year. This pioneering effort seeks to bridge the gap in data availability that has hampered major AI models in accurately reflecting Quebec’s rich heritage.
Addressing Data Deficiencies in AI
The primary motivation behind this initiative is to tackle the persistent issue of underrepresentation of Quebec-related data in AI training datasets. BAnQ has acknowledged that many generative AI systems struggle to provide accurate information about Quebec’s society, economy, and culture, largely due to the scarcity of locally relevant data. “All scenarios are a little bit on the table right now,” stated Valérie D’Amour, who headed the feasibility study. She emphasised the importance of engaging with cultural stakeholders, data owners, and providers throughout the process to validate the potential of the proposed databank.
Marie Grégoire, president and CEO of BAnQ, articulated the project’s vision: “Our goal is to ensure that AI systems better reflect Quebec society and culture.” She highlighted the necessity of integrating Quebec references into AI models, whether sourced from research institutions or the business sector.
Learning from Global Examples
The initiative draws inspiration from similar projects around the globe. In Sweden, for instance, large collections of Nordic-language texts have been compiled to bolster the development of generative AI models tailored for Scandinavian languages. BAnQ plans to initiate the process with its own collections before expanding to include data from external sources.
This initiative is rooted in recommendations from a report released by Quebec’s innovation council in 2024, which identified the lack of Quebec-specific data as a significant barrier to effective AI training. Destiny Tchéhouali, co-holder of a research chair focused on French-language AI and digital technologies, underscored the critical need for increased representation of Quebec culture in AI corpora. He warned of the risks of perpetuating linguistic and cultural biases, particularly regarding Indigenous communities, if these gaps remain unaddressed.
A New Framework for Content Protection
As BAnQ forges ahead with the development of this database, concerns regarding copyright and creator compensation have surfaced. Grégoire has reassured creators that the proposed platform could provide better protection than the current landscape, which she described as akin to “the Wild West.” She explained that the database could serve as a centralised mechanism to ensure creators are fairly compensated for their contributions.
However, some artists express trepidation about the long-term implications of their work being used in AI training systems. Maxime Harvey, a postdoctoral researcher at the National Institute of Scientific Research, noted, “The main criticism we hear in the field is that, even if artists earn income from it, they are still feeding the beast that will eventually be used to replace contracts they may lose because of AI.” This sentiment reflects the complex balance between leveraging technology and safeguarding artistic livelihoods.
Financial Outlook and Future Plans
BAnQ’s feasibility study outlines a budget of approximately £10.5 million over the next five years to bring this vision to fruition, encompassing both operational and capital costs. The project has already received £340,000 from the Quebec government for the feasibility study and an additional £750,000 to support the forthcoming 12-month experimental phase. Although the platform is envisioned to become operational by 2029, D’Amour indicated that the timeline will be reassessed following the experimental phase.
Why it Matters
The creation of this cultural and governmental databank is not just a technological advancement; it represents a significant step towards preserving and promoting Quebec’s unique identity in the digital age. As AI continues to evolve, ensuring that it accurately reflects the diversity and richness of Quebec’s society is crucial. This initiative has the potential to empower local creators, foster cultural understanding, and mitigate the risks associated with AI biases, ultimately shaping a more inclusive future for both technology and the arts.
