According to the Motion Foresight Monitor, United States President's Council of Advisors on Science and Technology Co-Chair David Sacks published a post in support of Palantir CEO Alex Karp's sharp criticism of the Frontiers AI Lab, stating that mainstream media dismissing his interview as a "meltdown" precisely highlighted that Karp hit the nail on the head. Sacks pointed out that true "AI security" in a corporate setting is not about abstract "alignment research" or a government-led certification system, but rather about being able to control one's own data, model weights, and computing power—thus preventing frontier labs from extracting a company's proprietary knowledge and turning it into the next product. He quoted Karp's words: "They want to own the means of production, not hand it over to someone else."

Sacks used the conflict between Figma and Anthropic as a typical example: Three days before Anthropic released Claude Design, its Chief Product Officer was still a board member of Figma, with Figma's founders stating that Anthropic "had not been honest with them." Subsequently, Figma's stock price plummeted while Anthropic's valuation soared. He further listed products such as Claude Science, Claude Security, Claude Legal, and Claude Code, pointing out that each time Anthropic directly entered vertical domains originally served by enterprises relying on its models, the pattern remained the same—"observe where value is being created and then dive in directly."

Sacks believes that labeling open-source models as "dangerous" is not about enterprises—retaining the choice of model layers and deciding who can use their core strengths is the true security bottom line for enterprises. Palantir previously collaborated with NVIDIA to deploy the Nemotron open AI model in a sovereign environment, serving the U.S. government and critical infrastructure clients, helping organizations locally train and deploy AI while fully controlling data and intellectual property.

Recently, Palantir CEO Alex Karp sharply criticized leading AI model companies in an interview with CNBC's "Squawk Box," stating that the sales approach of AI is "completely wrong." Karp emphasized that current enterprises are dissatisfied with OpenAI, Anthropic, and other "frontier labs," believing that they only pursue token maximizing, causing companies to waste time and money while handing over proprietary value and IP. Karp expressed that the enterprises are "angry" and will focus on autonomously owning AI means of production rather than relying on third parties.

On June 29, Palantir partnered with Nvidia to deploy the Nvidia NeMoTron Open AI Model in a sovereign environment, primarily serving the U.S. government and critical infrastructure customers. The collaboration reportedly integrates Nvidia's AI technology with Palantir's AIP, Foundry, Ontology, and Apollo platforms to help organizations locally train, customize, and deploy AI while fully controlling data, intellectual property, and models.

---------------------------------

Click the original text link below to join the DCSights Beating · FeiShu AI News Channel, monitoring global AI hot topics and news 24/7.