Microsoft’s multi-year partnership with Mistral AI has drawn the attention of European regulators amid concerns over the scope of the agreement between the two companies.
The partnership, officially announced this week, will give Mistral AI access to Microsoft’s Azure and cloud-based AI infrastructure to support the development and deployment of future Mistral large-scale language models (LLMs).
The deal is also an opportunity for Mistral AI to “explore new commercial opportunities” through expanded access to global markets, Microsoft said.
Microsoft reveals that the partnership will focus on three key areas, including allowing Mistral to leverage the tech giant’s Azure AI supercomputing infrastructure to enhance the scale and performance of its models; I made it.
Microsoft will also expand Mistral’s current offerings into the market, making its premium products available to customers in Azure AI Studio and the Azure Machine Learning Model Catalog, and allowing Mistral to “promote its models to Microsoft customers around the world.” , sell, and distribute.
The partnership will enable the two companies to experiment with testing and training purpose-built models for a diverse customer base, including “European public sector workloads,” and joint research and development.
Mistral AI chief executive Arthur Mensch hailed the partnership as a “major milestone” for the French startup. The company has established itself in recent months as a leading European competitor to US-based dominance in generative AI.
Last year, Mensch set out a goal to “create a European champion” that could compete with OpenAI and other major industry players.
But while the partnership with Microsoft may be a big stamp of approval for the company, EU regulators have expressed concerns about the scope of the partnership.
Union lawmakers appear to be taking a broad view of the announcement, and have already voiced concerns about Microsoft’s investment in the company.
A European Commission spokesperson said: “The European Commission is investigating agreements concluded between large digital market participants and generative AI developers and providers.” politiko. “In this regard, we have received the mentioned agreement and will analyze it.”
It was initially revealed that the tech giant had pledged $16 million in funding to the Paris-based startup. However, confusion regarding this arose after a Microsoft spokesperson said: Reuters The company reportedly invested in the company without owning any shares.
Microsoft later revealed that this investment would be converted into stock in an upcoming fundraising event.
Alex Hafner, a competition partner at British law firm Floodgate, said the key issue for regulators in this case is Microsoft’s “extent of minority interest” in Mistral.
He said: “EU merger control rules only have an impact where one party has gained ‘decisive influence’ over the other, which essentially means that through their rights as shareholders , or that the other party has the ability to influence the target’s business conduct through relevant contractual means.” .
Hafner added that lawmakers are keen to fully establish the relationship between the two companies post-contract.
“The first thing European legislators are likely to focus on is how the new deal will work in practice and, more specifically, how Microsoft will deal with Mistral in the light of the EU merger control framework/regulations. What kind of rights have you acquired?”
Microsoft is no stranger to regulatory scrutiny
Microsoft’s relationship with its partners in the generative AI space, particularly OpenAI, has been a source of contention in recent months.
In late 2023, competition regulators in the UK, US and EU launched preliminary investigations into the relationship between the two companies over concerns that the partnership could have a negative impact on competition in the artificial intelligence (AI) market.
A key factor in these investigations is the extent of Microsoft’s control over OpenAI activities. The UK’s Competition and Markets Authority (CMA) has raised concerns about whether Microsoft’s multibillion-dollar investment in OpenAI constitutes a “taking of control.”
Similar observations have been made by both the Federal Trade Commission (FTC) and EU merger regulators, both of whom have responded to the firing of OpenAI CEO Sam Altman and the subsequent addition of Microsoft personnel to the board. It noted that recent controversies raise further questions about the scope of control. .
