AI companies are pouring hundreds of billions of dollars into AI models, chips and data centers. But the increasingly intertwined financial relationships supporting that expansion could also amplify risks if AI demand falls short of expectations, according to a new Moody’s Ratings report.
Need more clues? Ask the Sherlock chatbot in the lower right corner to summarize this story, explain technical concepts or answer other questions.
“These tie-ups permeate the broader AI value chain, with the top AI labs representing a large share of the revenue backlog for the hyperscalers and significant direct and indirect customers of the semiconductor suppliers,” the report said.
“The partnership, ownership and guaranty agreements … magnify the already significant risk should the massive projected demand for AI products fail to materialize.”
“While strategic equity stakes in emerging technologies have been common among established tech companies, the degree of interdependence among AI labs, hyperscalers, semiconductor suppliers and sponsors is greater than traditional venture capital relationships, representing a more circular system that could mask true demand.”
Moody’s estimates that the backlog of contracts among Amazon, Microsoft, Alphabet and Oracle is $2.1 trillion as of their last reported quarter, with OpenAI and Anthropic taking up a large share of the contracted backlog.
The credit ratings agency cited the following circular relationship examples:
» Microsoft is a principal investor in OpenAI and provides it with cloud services. The two companies have revenue-sharing agreements, the details of which are unclear.
» AWS is the primary cloud provider for, and a large investor in, Anthropic.
» Nvidia agreed to acquire equity in OpenAI in exchange for OpenAI committing to using Nvidia’s systems to build AI data centers.
» AMD agreed to grant equity to OpenAI in exchange for OpenAI committing to deploy AMD’s
rack-scale systems for AI clusters. AMD also agreed to invest as much as $5 billion in Anthropic, with the AI lab agreeing to buy and deploy up to 2 gigawatts of AMD’s GPUs and new rack-scale systems.
Moody’s said these arrangements make it increasingly difficult to determine whether AI demand is entirely market-driven or partly sustained by strategic investments and long-term commitments among the industry’s biggest players.
If AI spending slows, the report warned, financial stress could spread through the ecosystem because so many companies now serve simultaneously as customers, suppliers, investors and financing partners.
Be First to Comment