Quebec’s Bibliothèque et Archives nationales du Québec (BAnQ) is taking significant strides towards developing a comprehensive database aimed at enriching artificial intelligence systems with cultural and governmental content reflective of the province’s diverse society. Following a detailed feasibility study, the library has embarked on an experimental phase to create a databank that will encompass both French and Indigenous languages, addressing the critical gap in Quebec-related data available for AI training.
Addressing Data Limitations
The impetus for this initiative stems from concerns that prominent generative AI models often fail to accurately represent Quebec’s unique culture, society, and economy due to a scarcity of relevant data. Valérie D’Amour, who spearheaded the feasibility study, expressed optimism about exploring various scenarios as the project unfolds. “All scenarios are a little bit on the table right now,” she noted. The collaborative effort will involve discussions with cultural stakeholders and data providers to validate potential avenues for development.
BAnQ has clarified that this new platform will not serve as an open distribution network for creative works; instead, access to the data will remain strictly regulated. Marie Grégoire, the president and CEO of BAnQ, emphasised the goal of ensuring that AI systems better mirror the realities of Quebec life, stating, “That means having Quebec references, whether in small models or large models, whether they come from research or from the business community.”
Learning from Global Initiatives
This initiative aligns with similar efforts observed internationally, such as in Sweden, where extensive collections of Nordic language texts have been curated to foster the development of AI models for Scandinavian languages. BAnQ plans to first utilise its own collections before expanding to include data from external sources, thereby establishing a robust foundation for the databank.
The recommendation to pursue this project was articulated in a 2024 report by Quebec’s innovation council, which highlighted the insufficient volume of Quebec-specific data in existing AI training datasets. Destiny Tchéhouali, a leading figure in Quebec’s research community focused on French-language AI and digital technologies, pointed out that the current AI landscape tends to overlook Quebec culture. He cautioned against the risks of perpetuating linguistic and cultural biases, particularly regarding Indigenous communities, stating, “We run an even greater risk of all these biases.”
Safeguarding Creative Works
As BAnQ develops its database, concerns surrounding copyright and the protection of creators’ rights have emerged within the cultural sector. Grégoire argued that the proposed platform could enhance protections for creators compared to the current landscape, which she likened to a “Wild West” scenario. She asserted, “Data is being harvested for free, and that should not be the case.” The initiative aims to create a centralised system that would facilitate fair compensation for creators whose works are utilised.
However, some artists remain apprehensive about the potential implications of contributing their creative outputs to AI training systems. Maxime Harvey, a postdoctoral researcher at the National Institute of Scientific Research, voiced a common concern among artists. “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,” he remarked.
Project Timeline and Funding
The feasibility study envisions the databank becoming operational by 2029, although D’Amour indicated that the timeline might be adjusted based on insights gained during the experimental phase. The projected budget for the next five years is estimated at nearly $10.5 million, which includes both operational and capital expenses. To support this ambitious undertaking, BAnQ has already secured $340,000 from the Quebec government for the initial feasibility study, along with an additional $750,000 to facilitate the 12-month experimentation phase.
Why it Matters
The establishment of this databank represents a crucial step towards ensuring that artificial intelligence systems can accurately reflect the rich tapestry of Quebec’s society and cultural heritage. By addressing the current disparities in data representation, BAnQ not only aims to safeguard the interests of local creators but also strives to foster a more inclusive and culturally aware AI landscape. As the project progresses, it could set a precedent for how cultural data is handled in the digital age, ensuring that Quebec’s unique voices are heard and respected in the evolving technological narrative.
