Monopoly of large AI models triggers "data sovereignty" crisis, intensifying conflicts between US technology giants and emerging start-ups
With the accelerated iteration of generative artificial intelligence (AI) technology, the conflict of interests between traditional American technology giants, real enterprises and emerging AI laboratories is becoming increasingly fierce. Many heads of large companies and industry analysts have spoken out recently, publicly accusing emerging AI unicorns such as OpenAI and Anthropic of over-extracting the value of customer data, charging high service fees, and trying to use the advantages of the underlying model to cannibalize the traditional software market. Market analysts believe that this game around "technical value distribution" and "enterprise data sovereignty" is reshaping the competitive landscape of the global technology industry. Alex Karp, CEO of Palantir Technology, a Silicon Valley big data analysis giant, recently publicly stated that strong dissatisfaction with emerging AI laboratories in the U.S. business community is erupting. Karp pointed out that a large number of enterprise-level customers did not obtain corresponding business value after paying high fees for computing power tokens (Tokens). In response to this current situation, Palantir released a white paper titled "Institutional Sovereignty in the AI Era" this week, formally proposing 15 specific defense recommendations, calling on companies and government agencies to take action to prevent their core business data and decision-making mechanisms from being "ransacked" by large model developers. This core pain point quickly sparked a structural debate across the tech industry. Former White House artificial intelligence advisor David Sacks analyzed this and pointed out that AI laboratories represented by Anthropic are exposing a very consistent monopoly path - that is, obtaining core business data by dominating the underlying large model, and then directly launching vertical applications covering science, security, law, programming and other fields, accurately eroding the market share of downstream partner companies. Unlike Palantir’s radical public opinion offensive, traditional software giants The company chooses to issue compliance warnings from the level of the enterprise value chain. Microsoft CEO Satya Nadella emphasized in a speech at Stanford University this month that if traditional companies only act as "consumers" of large models, they will not be able to retain their core corporate values, let alone create new value. He warned that companies must establish retained ownership of "technology-derived experiences" as they deploy AI. Market data also shows that due to concerns about the leakage of core business secrets and soaring operating costs, there have been reports such as Large traditional companies such as Microsoft and others have begun to try to use their own technology to replace Microsoft, For systems purchased by traditional software vendors, industrial shocks are spreading to the real economy. Faced with accusations of "excessive premiums" for large models, major giants are trying to break the monopoly of emerging laboratories through technology open source and price wars. Mark Zuckerberg, CEO of Meta Platform Company, made it clear in an interview with Bloomberg that the current pricing mechanism of some AI laboratories is too extreme and has the nature of huge profits. Meta plans to provide cutting-edge intelligent services at lower market costs by launching a new version of the paid tier of AI models, directly attacking the pricing weaknesses of competitors such as OpenAI. In addition, given the intertwined interests among technology giants, the policy winds in Washington have also changed. It is reported that in The Trump administration formally suspended export licenses for Anthropic's latest core large model last month after CEO Andy Jassy raised concerns about the safety of the technology's output. Industry analysts generally believe that as emerging AI laboratories such as OpenAI and Anthropic gradually advance their initial public offerings (IPOs), their high valuations and core control have aroused high alertness at the level of geopolitics and business competition. In this technological change that is regarded as "comparable to the invention of electricity", the offensive and defensive battle between traditional software giants and emerging AI forces has just established the boundaries, and there is still huge uncertainty in the future industrial structure.