The Federal Trade Commission and the European Commission for Competition are examining virtual worlds and generative artificial intelligence. This includes investigating partnerships like Microsoft’s relationship with OpenAI for possible antitrust violations.
Unfortunately, the regulators appear to ignore the reality that the success of AI technology depends on scale and resources. Moreover, these investigations represent an attack on one of the market’s least vertically integrated AI partnerships.
The investigation centers on two emerging technologies: virtual worlds and generative AI, which means AI that can create new content based on training data but without replicating the data used. Regulators are particularly interested in the partnership between Microsoft and OpenAI, the company behind ChatGPT, and whether that partnership threatens competition in the emerging AI market.
While not an acquisition, the partnership has developed amid a wave of large tech companies acquiring AI startups. This has prompted regulatory agencies, like the FTC, to raise concerns about whether these partnerships are a threat to competition.
The relationship between Microsoft and OpenAI isn’t as cut and dry as it may first appear because it is not a traditional acquisition. Microsoft offers OpenAI access to cloud computing and other resources necessary to run a large language model. Through this model, Microsoft has invested billions in OpenAI to help it develop an impossible scale.
The differences between Microsoft and Google regarding generative AI raise questions about whether regulators ignore the industry’s economic realities and hamper, rather than encourage, competition.
While testifying before the FTC, Microsoft argued that its partnership with OpenAI is necessary to compete in the AI market, specifically against Google.
Microsoft said that Google owns every stage of production required to create generative AI, such as microprocessors and massive amounts of data, through its ownership and operation of Google Search and YouTube.
Microsoft argues that this puts Google in a unique position to compete without requiring partnerships with AI developers. As the AI industry expands, different companies will try new approaches to compete with one another, whether that be through partnerships or creating the technology in-house.
Moreover, while it is certainly possible that a future breakthrough will allow smaller companies to create generative AI, right now, size matters.
Developing and training an AI model requires massive computing power to handle all the data necessary for providing consumers with quick responses to questions and prompts. This involves infrastructure that goes far beyond what a startup usually has access.
It is unclear why the partnership between Microsoft and OpenAI is subject to such regulatory scrutiny when it is the least integrated option available for developing AI at scale. While neither the Microsoft/OpenAI nor the Google model is bad, time will tell which one becomes more successful, if either.
As the industry continues to develop, we will see which business models work, but unless a discovery allows AI to become less complicated, large firms will have to be involved.
The FTC and the European Commission’s investigation into Microsoft and OpenAI’s relationship is a strange attack on a less concentrated approach to AI development. The United States and the European Union would be wise to reconsider its probe and allow market competition to develop free from government interference.
