Barely announced, the partnership between Mistral AI and Microsoft is already being called into question. He will be “examined” by the European Commission, declared a spokesperson for the institution while no official communication has yet been made. This commercial partnership involves “an investment of 15 million euros in Mistral AI“by Microsoft”which can be converted into capital during Mistral's next fundraising”, confirmed the American company and the French start-up with The Digital Factory.
Problem: the partnership concluded – under the guise of pushing the marketing of its model on Azure AI and helping to develop future dedicated models – above all locks access to the ecosystem developed by the start-up. Microsoft itself recognizes this: “It also makes deployment of premium Mistral AI models available to customers through our Models-as-a-Service (MaaS) offering on Microsoft Azure, which model developers can use to publish and monetize their AI models.”
Advertisement
Intense lobbying carried out in Brussels
This announcement, unsurprisingly, caused an outcry within the European institutions. Reuters thus reports that the agreement with Mistral raised eyebrows in Brussels, particularly due to pressure over the last year over the Artificial Intelligence Act (AI Act). As a reminder, France, alongside Germany andItalyhad urged Brussels not to over-regulate so as not to restrict innovation.
For its part, Mistral had pushed for exemptions for certain AI systems, warning that too strict legislation would hamper European start-ups' chances of competing with American giants. The deal between Mistral and Microsoft has therefore led some lawmakers to question the company's motives.
The words of Kai Zenner, head of office and digital policy advisor to Axel Voss, MEP of the European People's Party (EPP), interviewed by Euronews Nextshow a certain exasperation: “on a technical and political level in the (European) Parliament, we are extremely furious because the French government has put forward for months this argument of European leadership, meaning that these companies should be able to develop without the help of Chinese or American companies “.
MEPs step up to the plate
“This story appears to have been a front for the big US-influenced tech lobby”said Kim van Sparrentak, an MEP who worked closely on the AI Act. “The law almost collapsed under the guise of no rules for 'European champions', and now look. European regulators have been fooled”, she lamented. For her part, MEP Alexandra Geese judged that this announcement raised legitimate questions about the behavior of Mistral and the French government during the negotiations: “There is a concentration of money and power here like the world has never seen, and I think it deserves investigation.”
Advertisement
In her clumsy attempt to save face for the government and the start-up, the brand new Secretary of State for Digital, Marina Ferrarisaid to The gallery : “I find it great that a young French company is able to achieve a partnership of this magnitude with Microsoft. We must therefore be delighted with this agreement.” And to continue on the issue of open source, gradually abandoned by Mistral AI: “Mistral has said for a long time that this is not its only strategy, and that it will need to address different markets with different solutions.”
For its part, Mistral AI defends itself as best it can from this dependence on American technologies and investors. “Mistral AI's mission is to promote the adoption of generative AI by as many people as possible. The portability of our models across all distribution platforms (Azure and AWS, editor's note) will not only enable this adoption but will contribute to to guarantee the independence of our company”, said a spokesperson for the French company.
A high concentration of power
So is Microsoft playing with fire and will it end up burning its fingers? To date, the firm is the subject of an injunction by the Federal Trade Commission (FTC) alongside Alphabet, Amazon, Anthropic and OpenAI. The survey aims to obtain “a better internal understanding of these relationships and their impact on the competitive landscape”, says the FTC.
More recently, it is the Competition Authority which has taken a close interest in the subject. It intends to examine closely the strategies carried out by the different players in the generative AI market, the investments but also the risks that this generates, particularly in terms of access to data and a qualified workforce. She therefore launched a public consultation until March 22.
Microsoft congratulates the legislative work carried out…
On the occasion of the Mobile World Congress (MWC) which is being held in Barcelona from February 26 to 29, Brad Smith, president of Microsoft, spoke about the work carried out by the Redmond firm in terms of support for innovation for AI. “Competitive pressure is fierce and the pace of innovation is dizzying. As a leading cloud provider and innovator in AI models, and through our partnership with OpenAI, we understand our role and responsibilities in shaping this era of AI” did he declare.
He continued by highlighting the legislative work carried out on this technology: “The European Union is implementing its Digital Markets Act and finalizing its AI Act, while the United States is moving quickly with a new AI Decree. Similar laws and initiatives are progressing in the United Kingdom , in Canada, Japanin India and in many other countries.
…And defends its strategy
And while innovation and competition will require a broad range of similar support for proprietary and open source AI models, large and small, Microsoft argues against favoring only Mistral AI: “we have also invested in a wide range of other generative AI startups“. He thus specifies that “in some cases, these investments have provided seed funding to fund day-to-day operations. In other cases, these investments have been more focused on paying for expenses related to using the IT infrastructure needed for training and the deployment of generative AI models and applications.
Selected for you