Conference Presentation, Fireside Chat, Panel
Who Controls AI? The Open Source Imperative | Arthur Mensch, Mistral AI | RAISE Summit 2026
- Economic risks identified include widening inequality where non-AI sectors effectively pay rent for AI growth, potentially leading to social unrest if value and intellectual property (IP) are not shared more broadly.
- Arthur frames AI strategy through three pillars analogous to the energy sector:
- Security of supply: Nations must avoid over-dependence on foreign imports by building domestic compute infrastructure and services decoupled from foreign interests.
- Affordability: Control over technology generation and distribution is linked to cost management.
- Sustainability: AI's environmental footprint necessitates careful long-term planning.
- Marc notes that while Canada and the US share these strategic concerns, European nations are currently leading the conversation on sovereignty.
- Sovereignty is redefined not as isolation, but as strategic collaboration with trusted partners to manage economic dependencies.
- Open source is identified as a critical lever for achieving national sovereignty because:
- It allows nations to "fork" and adapt technology, preventing reliance on external providers who could cut off access ("shut the lights off").
- It enables countries to build ownership without necessarily developing all components from scratch.
- Recent European frameworks and the Canadian AI strategy are beginning to recognize this role for open source.
- Arthur observes a significant shift in market sentiment over the last three years, where stakeholders moved from viewing AI as a simple "chatbot" to recognizing it as essential infrastructure for operating the global economy.
- Forward-looking consensus projects that within 5 to 10 years, open source components will become the dominant "Lego blocks" and building blocks of the AI landscape.
- While the trend toward open source dominance is considered highly probable, Marc and Arthur emphasize that momentum is not inevitability; success requires sustained R&D, funding, and interoperability between different open source stacks.
- Current challenges for the ecosystem include financing the massive capital investments needed to build advanced models and fostering partnerships among multiple companies with aligned interests.