In a significant move to better represent Quebec’s rich cultural tapestry in artificial intelligence systems, the Bibliothèque et Archives nationales du Québec (BAnQ) has embarked on an ambitious project to create a comprehensive database of cultural and governmental content. This initiative aims to improve AI’s engagement with Quebec society, including its diverse Indigenous languages, addressing the current shortfall of relevant data and fostering a deeper understanding of local contexts.
A New Approach to AI Data Representation
Earlier this year, BAnQ completed a feasibility study that laid the groundwork for the development of this innovative databank. Following the study, they have commenced an experimental phase focused on gathering data in both French and Indigenous languages. The project is particularly timely, as concerns have been raised that existing generative AI models often lack accurate and extensive information about Quebec’s unique social, cultural, and economic landscape.
Valérie D’Amour, who spearheaded the feasibility study, noted, “All scenarios are a little bit on the table right now. We have a lot of ideas and we want to validate the possibilities with cultural stakeholders, as well as with data owners and providers, who will be involved in the discussions.” This collaborative approach aims to harness the insights of various community members and experts to ensure the databank serves its intended purpose effectively.
Controlled Access and Cultural Integrity
BAnQ emphasises that the new platform will not serve as a public distribution channel for creative works. Access to the data will be stringently regulated. Marie Grégoire, BAnQ’s president and CEO, articulated the vision behind the initiative, stating, “The goal is to ensure that AI systems better reflect Quebec society and culture. That means having Quebec references, whether in small models or large models, whether they come from research or from the business community.”
This careful management of data is crucial, especially given the global context where AI models have often been built on datasets that overlook regional nuances. Grégoire highlighted the importance of a robust framework that protects local creators while also enriching the AI landscape with culturally relevant content.
Learning from Global Initiatives
Quebec is not alone in this endeavour; similar projects are underway in other countries. For instance, Sweden has been assembling large collections of Nordic-language texts to enhance AI capabilities in Scandinavian languages. BAnQ plans to initiate its project by utilising its own collections before expanding to other data sources. This strategy aligns with recommendations from Quebec’s innovation council, which identified the scarcity of local data as a significant barrier to effective AI training.
Destiny Tchéhouali, a prominent researcher focusing on French-language AI and digital technologies, remarked on the cultural implications of this initiative. He asserted that Quebec’s cultural identity is often inadequately represented in global AI datasets. “We run the risk of reproducing linguistic biases and cultural biases. When we also talk about Indigenous peoples, we run an even greater risk of all these biases,” Tchéhouali cautioned, underscoring the need for strategic infrastructure to catalogue and track local content.
Balancing Innovation and Creator Rights
As BAnQ moves forward with its databank, concerns regarding copyright and the potential exploitation of creators’ works loom large. Grégoire is adamant that the proposed platform could offer creators enhanced protection compared to the current landscape, which she described as akin to “the Wild West.” She argued that a centralized system would facilitate fair compensation for artists whose works are utilised in AI training.
Despite these assurances, some artists express trepidation about the implications of contributing to AI systems. Maxime Harvey, a postdoctoral researcher, voiced a common concern among creators: “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 a broader anxiety within the artistic community about the sustainability of their livelihoods in an increasingly automated world.
The feasibility study envisions the platform becoming operational by 2029, with a projected budget of nearly £8.5 million over five years, encompassing both operational and capital expenses. BAnQ has already secured funding from the Quebec government, receiving £340,000 for the feasibility study and an additional £750,000 to support the 12-month experimental phase.
Why it Matters
The establishment of this databank is not merely an academic exercise; it stands to redefine how artificial intelligence interfaces with Quebec’s cultural identity. By prioritising local knowledge and languages, BAnQ is taking a vital step toward ensuring that AI systems do not perpetuate existing biases but rather reflect the diversity and richness of Quebec society. This initiative could serve as a model for other regions grappling with similar challenges, highlighting the importance of local context in the global AI discourse. In doing so, it not only protects cultural heritage but also enhances the sustainability of creative industries in the face of rapid technological advancement.
