Meta、微软和沃尔玛等大型企业正在重新评估并放缓其快速采用人工智能工具的步伐，原因是出现了意外的运营后果。
Corporations Scale Back Rapid AI Deployments Amid Productivity Shortfalls and Operational Risks Major corporations including Meta, Microsoft, and Walmart are reassessing and scaling back their rapid adoption of artificial intelligence tools following unexpected operational fallout.
尽管已投入数十亿美元，但亚特兰大联邦储备银行的一份工作论文发现，90%的高管报告称未实现生产力提升，而公司正面临错误增加、安全风险上升和工作流程瓶颈等问题。
Despite billions invested, an Atlanta Federal Reserve working paper found that 90% of executives report no productivity gains, while companies confront mounting errors, rising security risks, and workflow bottlenecks.
Meta在发现员工花费过多时间修复AI错误后，正在重新提拔中层管理者；微软和麦肯锡则限制了员工使用AI，以控制成本和保护专有数据。
Meta is repromoting middle managers after discovering workers spent excessive time fixing AI errors, while Microsoft and McKinsey have restricted employee usage to contain costs and protect proprietary data.
在法律程序中，可疑的生成式AI输出导致3M被判承担6150万美元的责任。
In legal proceedings, questionable generative AI outputs contributed to a $61.5 million liability verdict against 3M.
匹兹堡大学的研究表明，由AI驱动的裁员未能兑现承诺的绩效改善，促使领导层放缓推广计划，并转向有针对性的员工培训。
Research from the University of Pittsburgh indicates that AI-driven layoffs have failed to deliver promised performance improvements, prompting leadership to slow rollout schedules and shift toward targeted workforce training.