Linux Foundation Research has published The Value of Open Source AI for the Canadian Economy, a review of industry, academic and open source evidence written by Hilary Carter and Anna Hermansen and commissioned by Meta. The report forecasts that AI could add up to 9% to Canada’s GDP by 2035 and $180 billion annually by 2030, and argues that the country’s open-by-default policy, open-weights releases and existing open source projects give it a strong foundation for adoption.
Taskd appears in the section on how Canadian startups use open models to keep build costs low and speed to market high. The report describes Ontario-based Taskd as using Meta’s Llama 3.2 to create, deploy and scale intelligent AI agents from fragmented data, code and workflows, while controlling for regulatory compliance. It cites Taskd’s maritime RFQ extraction demonstration, published on GitHub, as an example.
Llama models let a small team in Ottawa deliver enterprise results, turning unstructured information into a dynamic, private knowledge graph. Faster quotes, fewer errors, less waste. The data stays with the customer. Every step is cited and auditable, and the graph learns each cycle.
Why open models suit the work
Taskd’s architecture keeps a business’s operational context, rules, approvals and history separate from whichever model reads its documents. That separation is what makes open weights practical for regulated operations. The model can run inside the customer’s boundary, the data does not leave, and because the model is used to read rather than to decide, a change of model does not change what the business knows or how its rules behave.
The report was released alongside Meta’s announcement on 9 February 2026 of its findings on AI-driven economic and job growth in Canada. The full PDF is available from the Linux Foundation.