The Pentagon has replaced Anthropic's AI models with alternatives from xAI, OpenAI, and Google, according to Palantir's chief technology officer, Shyam Sankar. The change follows a dispute over Anthropic's AI safeguards, with no reported operational impact.
TL;DR
- The Pentagon has replaced Anthropic's AI models with alternatives from xAI, OpenAI, and Google.
- Palantir's Shyam Sankar stated that the switch had no operational impact.
- Sankar criticized effective altruism, claiming it fuels fear around AI, and discussed AI's role in military targeting.
What happened
Palantir's chief technology officer, Shyam Sankar, revealed on CBS's Face the Nation that the Pentagon has replaced Anthropic's AI models. The change occurred after a dispute over Anthropic's safeguards for military AI, with the Department of Defense labeling the company a supply chain risk.
According to Sankar, the Pentagon now uses models from xAI, OpenAI, and Google, as well as open-weight models like Nvidia's Nemotron. He emphasized that users building AI applications can choose and switch models, driving Palantir's commercial growth.
Why it matters
This shift highlights the Pentagon's flexibility in adopting AI models and its willingness to replace providers based on safeguard disputes. It also underscores the competitive landscape among AI companies vying for government contracts.
Sankar's comments on effective altruism and AI safety provide insight into the ongoing debate about AI's potential risks and the influence of certain ideologies within the AI research community.
Key facts
- The Pentagon replaced Anthropic's AI models due to a dispute over safeguards.
- Models from xAI, OpenAI, and Google, along with Nvidia's Nemotron, are now in use.
- Shyam Sankar stated that the switch had no operational impact.
- Sankar criticized effective altruism, comparing it to a 'doomsday cult'.
- He mentioned that many AI researchers have started smoking as a way to signal their beliefs about AI.
- Sankar described military targeting as a process, not a choice made by AI.
- Palantir has briefed Congress on Maven, its AI system used by the Pentagon.
Context
The replacement of Anthropic's models comes amid growing scrutiny of AI's role in military applications. The incident involving the US strike on a school in Minab, Iran, has raised questions about the reliability of AI in targeting decisions.
Sankar's comments on effective altruism reflect a broader debate within the AI community about the balance between innovation and safety. The Pentagon's ability to switch AI providers demonstrates the dynamic nature of the AI market and the importance of safeguards in military applications.
