帕兰蒂尔陷入了 Anthropic 与美国国防部的冲突之中，其首席技术官夏姆·桑卡尔在接受本·汤普森为《Stratechery》进行的采访时表示。该采访于周四发布。
Palantir was caught in Anthropic’s clash with the US Department of War, its chief technology officer, Shyam Sankar, told Ben Thompson for Stratechery, in an interview published on Thursday.
汤普森指出，Anthropic 为美国国家安全体系的工作是通过帕兰蒂尔进行的，并询问桑卡尔是否后悔。
Thompson noted that Anthropic’s work for the US national security system ran through Palantir, and asked whether Sankar regrets it.
“是的，我们被夹在中间。我不后悔。我们仍在与 Anthropic 合作，也与两家前沿实验室合作。”帕兰蒂尔首席技术官夏姆·桑卡尔对《Stratechery》表示。他说，许多当事方可能都没有妥善处理这场争端。自那以后，帕兰蒂尔投入了大量工作，让客户能够安全地将一种 AI 模型替换为另一种，他说。去年 10 月，桑卡尔曾告诉哥伦比亚广播公司（CBS），五角大楼已经替换掉了 Anthropic 的 AI。
“Yes, we were stuck in the middle. I don’t regret it. We still work with Anthropic, we work with both frontier labs.”Shyam Sankar, Palantir CTO, to Stratechery He said many of the parties probably did not handle the dispute well. Since then, Palantir has put a lot of work into letting customers swap one AI model for another safely, he said. In October, Sankar told CBS that the Pentagon had already replaced Anthropic’s AI.
关于 AI 安全 桑卡尔在采访前曾向汤普森发送了一份话题清单。其中一项询问有效利他主义是否已将 AI 安全变成了一种世俗宗教。“显然我认为是的，”他回答。
On AI safety Sankar had sent Thompson a list of topics before the interview. One asked whether effective altruism has turned AI safety into a secular religion. “Clearly I think so,” he answered.
他说，关于 AI 将带来乌托邦或人类灭亡的言论否定了人的能动性，连同随之而来的审慎设计 AI 系统的责任一并否定了。他说，他不介意任何人的信仰。他担心的是当这些信仰被作为技术性批评呈现给公众时。
He said talk of AI bringing either utopia or the end of humanity rejects human agency, and with it the duty to engineer AI systems carefully. He does not mind anyone’s beliefs, he said. His concern is when those beliefs are presented to the public as technical critiques.
关于客户与中国 桑卡尔表示，他见过的几乎所有主要利用 AI 裁员降本的客户，甚至连这一目标都未能达到。他说，成功的客户想利用 AI 赢得市场。
On customers and China Sankar said almost all the customers he has seen use AI mainly to cut staff and costs failed to reach even that goal. The ones succeeding want to use AI to win their markets, he said.
他说，帕兰蒂尔自身不使用 DeepSeek，但允许商业客户自行选择模型。他说，中国实验室目前仍落后于前沿水平。他预计，一旦中国实验室感觉自己处于前沿，就会停止发布开放权重模型。此后 DeepSeek 已发布其 Super-Agent 的代码。
Palantir does not use DeepSeek itself, but lets commercial customers choose their models, he said. He said Chinese labs are behind the frontier for now. He expects them to stop releasing open-weight models once they feel they are at the frontier. DeepSeek has since released the code for its Super-Agent.
汤普森在采访中反驳道。他说，桑卡尔的论点符合帕兰蒂尔的商业模式，正如Anthropic的安全警示符合其自身模式一样。桑卡尔回应称，两家公司均由创始人领导，且各自的战略都建立在其信念之上。Anthropic曾告知投资者，其安全立场可能使其成为政治攻击目标。
Thompson pushed back during the interview. He said Sankar’s arguments fit Palantir’s business model, just as Anthropic’s safety warnings fit its own. Sankar replied that founders lead both companies, and that each builds its strategy on what it believes. Anthropic has told investors its safety views could make it a political target.