Quebec’s national library is embarking on a significant project to establish a database aimed at enriching artificial intelligence systems with a deeper understanding of the province’s diverse culture, society, and Indigenous languages. The Bibliothèque et Archives nationales du Québec (BAnQ) has commenced the initial phase of this government and cultural databank in both French and Indigenous languages, following a feasibility study completed earlier this year. This initiative is a response to the pressing need for more reliable and relevant data about Quebec in the AI landscape, which has often been hampered by insufficient regional information.
Addressing the Data Gap
“Currently, AI models struggle with accurately reflecting Quebec’s distinct culture and societal nuances,” remarked Valérie D’Amour, who spearheaded the feasibility study. The project aims to bridge this gap by validating various possibilities with cultural stakeholders, data providers, and other involved parties. D’Amour emphasised that “all scenarios are a little bit on the table right now,” indicating a collaborative approach to developing the databank.
The proposed platform is not intended to serve as a public distribution channel for creative works; rather, access to the data will be meticulously regulated. Marie Grégoire, BAnQ’s president and CEO, highlighted that the primary objective is to ensure AI systems can better mirror Quebec’s rich cultural landscape. She stated, “This means incorporating Quebec references, whether they come from research or the business community, in both small and large models.”
Learning from Global Initiatives
This initiative aligns with similar efforts globally, such as Sweden’s compilation of Nordic-language texts to enhance generative AI models for Scandinavian languages. BAnQ plans to leverage its own collections as a starting point before branching out to include data from external sources.
The groundwork for this initiative stems from a recommendation in a 2024 report from Quebec’s innovation council, which pointed out the “very small quantity of data on Quebec” currently found in AI training datasets. Destiny Tchéhouali, a co-holder of a Quebec-based research chair dedicated to French-language artificial intelligence, expressed concerns over the underrepresentation of Quebec culture in prevalent AI corpora. “We risk reproducing both linguistic and cultural biases,” Tchéhouali cautioned, particularly regarding Indigenous communities.
Protecting Creators’ Rights
As BAnQ advances its databank project, copyright concerns have surfaced, raising questions within the cultural sector about the implications of AI on artists’ livelihoods. Grégoire argues that the new platform could provide artists with more robust protections compared to the current landscape, which she likened to “the Wild West.” She insisted that data should not be harvested freely and posited that the databank could act as a centralised gateway to ensure that creators receive due compensation for their contributions.
However, apprehension persists among some artists. Maxime Harvey, a postdoctoral researcher and member of the aforementioned research chair, noted that there is considerable unease about the potential for AI to replace traditional contracts. “Even if artists earn income from this, they may ultimately be feeding a system that could jeopardise their future opportunities,” he cautioned.
Looking Ahead
The feasibility study suggests that the databank could become operational by 2029, although D’Amour indicated that the timeline would be reassessed after the experimental phase. The estimated budget for the next five years stands at nearly £8.6 million, covering both operational and capital expenses. To date, BAnQ has secured £267,000 from the Quebec government for the feasibility study and an additional £600,000 to support the project’s year-long experimentation phase.
Why it Matters
This initiative is pivotal for Quebec as it seeks to assert its cultural identity in the rapidly evolving realm of artificial intelligence. By enhancing the representation of Quebec’s unique societal fabric and Indigenous languages in AI training datasets, the project not only aims to create a more accurate reflection of the province but also strives to empower local creators. In a world where AI increasingly shapes cultural narratives, ensuring that Quebec’s voices are heard and respected is essential for fostering a sustainable cultural ecosystem.
