Palantir CEO Alex Karp has renewed his criticism of leading AI companies, arguing that many of the industry’s largest model developers cannot be trusted with enterprise customers’ data.
In his annual letter to shareholders released Monday alongside Palantir’s latest earnings report, Karp—who holds a PhD in social theory and has long blended philosophy with business commentary—claimed that some AI companies are pursuing strategies that resemble ideas traditionally associated with Marxism.
“There are Marxist overtones and undertones to our business,” Karp wrote. “Others, including many of those building large language models, intend, knowingly or otherwise, to capture the means of production of their purported partners.”
Despite his criticism of AI model providers, Palantir has been one of the biggest beneficiaries of the rapid expansion of artificial intelligence. The company posted another record-breaking quarter, reporting $1.9 billion in revenue, a 93% increase compared with the same period a year earlier, along with $1.1 billion in profit. Karp noted that the company’s quarterly profit exceeded its total revenue from the same quarter last year.
During Palantir’s earnings call with Wall Street analysts, Karp expanded on his comparison using the patriotic, defense-focused rhetoric that has become a hallmark of the company. Palantir, whose executive leadership team is entirely male, has long positioned itself as a technology provider focused on government and national security customers.
Karp questioned whether businesses wanted to support what he described as a future in which companies help their competitors gain an advantage while only a small group ultimately benefits. He argued that a handful of technology firms believe they deserve control over the country’s “means of production,” while everyone else bears the cost of that transformation.
Palantir presents itself as an alternative by offering AI software that is model-agnostic, allowing governments and enterprises to choose different AI models while retaining ownership of their data, prompts, workflows, and operational context—what the company often refers to as AI “exhaust.”
Continuing his criticism, Karp argued that businesses are effectively paying AI providers to absorb their intellectual property and expertise. He suggested that organizations are funding the development of models that could eventually compete against them, claiming some AI companies believe they have the moral authority to take control of enterprise knowledge and capabilities.
Although Karp’s language was highly provocative, the broader concern he raised has become increasingly common across the technology industry. Microsoft CEO Satya Nadella has also cautioned enterprises against becoming overly dependent on a single AI provider or surrendering control of their proprietary data.
Supporters of this argument point to several examples where businesses partnered with or paid companies such as Anthropic and OpenAI, only to later see those same AI firms expand into markets including design software, healthcare operations, legal services, and drug discovery.
Ultimately, however, the rapidly evolving AI market remains large enough for multiple companies to succeed. Rather than producing clear-cut heroes or villains, the sector continues to expand at a pace that leaves room for both AI model developers and infrastructure providers like Palantir—a trend reflected in the company’s latest financial results.
