David Sacks supports Palantir CEO's criticism of AI lab: True enterprise AI security is about controlling one's own data, models, and computing power According to Beating Monitoring, David Sacks, co chair of the US President's Council of Advisors on Science and Technology, has issued a statement in support of Palantir CEO Alex Karp's sharp criticism of cutting-edge AI labs, stating that mainstream media's downgrade of his interviews as a "catastrophic loss of composure" precisely indicates that Karp has hit the nail on the head. Sacks pointed out that the true "AI security" in the enterprise environment is not abstract "alignment research" or government led certification systems, but whether one can control their own data, model weights, and computing power - so that cutting-edge laboratories cannot "suck up" the proprietary knowledge of the enterprise and turn it into the next product. He quoted Karp's words: "They want to own the means of production themselves, not hand them over to others." Sacks used the conflict between Figma and Anthropic as a typical case: Three days before releasing Claude Design, Anthropic's Chief Product Officer was still a member of Figma's board of directors, and the founder of Figma stated that Anthropic "did not always remain honest with them; Afterwards, 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 Anthropic always directly enters the vertical fields served by enterprises that rely on its models, with a consistent pattern of "first observing where value is created, and then directly entering". Sacks believes that the condemnation of open source models as "dangerous" is not for enterprises - retaining the choice of model layers and deciding who can use their core advantages is the true security bottom line for enterprises. Previously, Palantir had collaborated with Nvidia to deploy Nemotron's open AI model in a sovereign environment, serving the US government and critical infrastructure clients, helping institutions fully control data and intellectual property while training and deploying AI locally. Palantir CEO Alex Karp recently harshly criticized cutting-edge AI modeling companies in an interview with CNBC's "Squawk Box", stating that the sales approach of AI is "completely wrong". Karp emphasized that companies are currently dissatisfied with "cutting-edge laboratories" such as OpenAI and Anthropic, believing that they only pursue token maximization, wasting time and money while giving away proprietary value and IP. Karp stated that the company is' angry 'and will strive to independently own AI production materials rather than relying on third parties. On June 29th, Palantir partnered with Nvidia to deploy Nvidia Nemotron's open AI model in a sovereign environment, primarily serving the US government and critical infrastructure clients. It is reported that the collaborative system integrates Nvidia AI technology with Palantir AIP, Foundry, Ontology, and Apollo platforms to help institutions train, customize, and deploy AI locally, while fully controlling data, intellectual property, and models. ---------------------------------Click on the original link below to join the Beating · Feishu AI news channel, which continuously monitors global AI hotspots and news 24/7. [Original link]