微軟向OpenAI開炮：內部銷售備忘錄為何抨擊昔日合作夥伴
微軟正培訓銷售人員將openai和anthropic與其自研模型進行負面比較，聲稱性能更佳且更安全。這標誌著微軟為了降低成本，正推動自研ai並從合作轉向直接競爭。
microsoft is reportedly training its sales force to compete aggressively against openai and anthropic by criticizing their models compared to microsofts in-house phi models. evp jacob andreou presented data claiming anthropics claude was slower and less accurate than copilot within office apps and lacked proper security integrations. this pivot comes after microsoft recently removed the exclusivity clause from its openai partnership and began replacing external models with proprietary ones in flagship products like word and excel. the shift aims to reassure investors concerned about microsofts massive ai spending by demonstrating the competitiveness and cost-efficiency of its own technology.