Anthropic 的 IPO 會因自身的道德準則而失敗嗎
anthropic 提交了 IPO 申請以支持其 9650 億美元的估值，但其拒絕讓軍方使用其 AI 模型的立場引發了制裁戰，可能導致數十億美元損失。
anthropic confidentially submitted paperwork to us regulators for an ipo that could rival the largest ever, aiming to secure capital for expensive ai infrastructure. the company reported $47b in annualized revenue but remains unprofitable due to high cloud compute and staffing costs. this financial push collides with a major political risk: former defense secretary pete hegseth sanctioned anthropic, barring its claude models from federal agencies over ethical resistance to weapons deployment. the company is suing to overturn these sanctions, arguing its safety stance shouldn't preclude government work, setting up a high-stakes test of whether ai safety principles can survive public market demands.