In a landmark initiative aimed at enhancing the representation of Quebec culture in artificial intelligence systems, the Bibliothèque et Archives nationales du Québec (BAnQ) has announced the launch of an experimental database. This project seeks to compile cultural and governmental content in French and Indigenous languages, addressing the growing concern about the lack of reliable data regarding Quebec society available for AI training. Following a thorough feasibility study completed earlier this year, BAnQ is now moving forward to ensure that AI systems can better reflect the province’s diverse cultural landscape.
A Step Towards Cultural Representation in AI
The BAnQ’s plans stem from recommendations made by Quebec’s innovation council in 2024, which highlighted a glaring gap in the quantity of Quebec-specific data accessible for AI training purposes. “All scenarios are a little bit on the table right now,” remarked Valérie D’Amour, who spearheaded the feasibility study. The aim is to engage with cultural stakeholders and data providers to validate the possibilities for this innovative databank.
Marie Grégoire, BAnQ’s president and CEO, emphasised the importance of integrating Quebec references into AI models, whether from academic research or the business sector. “We want to ensure that AI systems reflect the reality of Quebec society and culture,” she stated. The initiative echoes similar efforts in countries like Sweden, where collections of Nordic-language texts have been assembled to bolster generative AI models for Scandinavian languages.
Control and Access: A Balancing Act
While the proposed database promises to enhance AI’s understanding of Quebec culture, BAnQ has made it clear that the platform will not serve as a public distribution channel for creative works. Access to the data will be meticulously regulated. Grégoire believes that this approach can lead to improved protections for creators compared to the current system, which she described as akin to “the Wild West.”
The database is envisioned as a centralised resource that would facilitate the compensation of artists whose works may be used in AI training. “By collaborating, cultural organisations can ensure that creators are fairly paid and that the sector remains sustainable,” Grégoire added. However, concerns persist among some artists regarding the potential impact of AI on their livelihoods.
Maxime Harvey, a postdoctoral researcher at the National Institute of Scientific Research, voiced apprehensions that contributing their work to AI training systems could jeopardise contracts for artists in the future. “Even if artists earn income from it, they could be feeding a system that ultimately replaces their jobs,” he cautioned.
Project Timeline and Financial Support
The feasibility study outlines a timeline for the database to become operational by 2029, although D’Amour indicated that this schedule may be reassessed post-experimental phase. The estimated budget for the project spans nearly $10.5 million through 2030, covering both operational and capital expenses. To support this endeavour, BAnQ has secured $340,000 from the Quebec government for the feasibility study, along with an additional $750,000 to fund the upcoming 12-month experimental phase.
Why it Matters
The development of this cultural databank is crucial not only for elevating Quebec’s unique voices within the realm of artificial intelligence but also for safeguarding the interests of local artists and creators. As AI continues to evolve and permeate various sectors, ensuring that diverse cultural perspectives are included in its training datasets is vital. This initiative represents a proactive approach to combatting the underrepresentation of Quebec culture in AI, ultimately fostering a more equitable and diverse digital landscape. The success of this project could set a precedent for similar efforts worldwide, reinforcing the importance of cultural integrity in technology.
