跳到正文
TNW· Ana-Maria Stanciuc·· 27 天前AI 評分74

Anthropic 在 IPO 文件中警告,AI 安全立場或令其成為政治攻擊目標

Anthropic tells IPO investors it could be targeted over its AI safety views

AI 導讀

Anthropic 在 IPO 的 S-1 註冊聲明中警告,公司的 AI 安全立場、政策倡議及高層言論可能招致政治批評、監管審查或不利對待。文件列出調查、選擇性執法、政府合約及員工簽證決策等風險,並稱公司預期繼續就 AI 政策發聲。文件亦提及與美國政府的多項爭議、台灣晶片製造集中及新資料中心面對的政治反對等風險。

正文

Anthropic has warned would-be investors that its views on AI safety could make it a political target. The warning is in the risk factors of its S-1 registration statement. The company made it public on Monday, setting out its plans for the Anthropic IPO on Nasdaq.

The filing does not name President Donald Trump or his administration. Owen Dahlkamp and Sam Sutton of Politico quoted the key passage:

“Our public positions, stated values, policy advocacy, or statements by our executives have attracted, and may continue to attract, criticism from influential political figures,” the company wrote.

Anthropic listed what that could mean. It might face investigations, or see rules applied to it selectively. It could also be treated worse in government contracts, or in visa decisions for its staff, according to Politico.

The risk of being singled out

Maria Curi of Axios quoted another line from the same section:

“The political environment and our engagement with policymakers and public positions on AI regulation and safety could subject us to heightened scrutiny, political backlash, or disproportionate regulatory action,” the filing states.

Anthropic also says it could face enforcement that rivals in a similar position might not, Axios reported. It does not promise to change course. It says it has backed political groups and spoken out on AI policy, and expects to keep doing so.

The filing lists its open fights with Washington. One is the court case over the Pentagon’s supply chain risk label. Anthropic has asked for a full-court rehearing there. Others are the risk that Commerce Department export controls return, and a Federal Trade Commission inquiry. It also cites the White House’s pre-deployment framework for AI models. Its deals with Amazon, Google, Microsoft and Broadcom could draw antitrust scrutiny, it says.

Trump’s “perfect little angel” post

Trump went after Anthropic’s chief executive in a Truth Social post in September, Politico noted. He was arguing against more AI rules.

“The Trump Administration has stopped AI ‘people’ from doing bad, or potentially bad, ‘things,’ like Dario (Anthropic!), who is now pretending to be a ‘perfect little angel’ – and we will continue to do so,” Trump wrote.

Amodei went to a White House dinner that month. Anthropic has also sent co-founder Tom Brown to work with Trump officials, Axios reported.

Taiwan and data centres

The filing names other risks beyond politics. Anthropic pointed to Taiwan’s dominance in chip manufacturing, which it said is especially prone to instability. Tension in the region, including threats from China, or natural disasters could disrupt supply, according to Politico.

“Efforts to establish sufficient advanced chip manufacturing capacity outside of Taiwan, including in the United States, remain in early stages,” the filing reads.

It also flagged political opposition to new data centres, which it said has grown and become more organised. Losing a major site might not be easy to make up with its other facilities, the company added.

來源:TNW · thenextweb.com