OpenAI and Microsoft Revise Deal for Future IPO Plan: Report

TheCryptoTimes发布于2025-05-12更新于2025-05-12

OpenAI and Microsoft are reportedly reshaping their long-standing partnership in a move that could clear the path for OpenAI’s initial public offering (IPO), according to the Financial Times. The talks aim to balance Microsoft’s deep investment in the ChatGPT maker with OpenAI’s ambitions to restructure its business model.

Microsoft, which invested an initial $1 billion in OpenAI in 2019 and later added over $13 billion, is now negotiating new terms. The software giant is said to be willing to give up part of its equity stake in OpenAI’s for-profit unit in exchange for long-term access to future AI models developed after 2030, when their current agreement expires, according to an FT report.

The original deal between the two was built when OpenAI was still transitioning from a nonprofit to a capped-profit structure. However, after pushback from former employees, researchers, and even rivals like Elon Musk, OpenAI is reportedly reconsidering its shift to a more traditional business model.

The changes at hand are considered instrumental for OpenAI as it moves into its next phase and contemplates eventual public listing. Having invested billions, Microsoft wants to protect its stake as OpenAI continues to make strategic changes.

Although the matter remains silent from both sides, insiders believe that the talks are underway and might fundamentally change the nature of OpenAI’s collaboration with Big Tech in the future.

The changes to come suggest that OpenAI is about to enter an innovative phase characterized by greater transparency and public market involvement.

Also Read: Mastercard Partners with Microsoft to Let AI Shop and Pay for You



你可能也喜欢

ACL 2026华人霸榜,最佳论文一作全华人,杰出论文几乎包场

ACL 2026计算语言学顶会于7月在美国圣地亚哥举行,规模创历史新高,共收到12148篇投稿,同比增长45%。本届会议被大语言模型(LLM)主题主导,相关词汇在论文标题中出现频率极高。 会议评选出三篇最佳论文,一作均为华人学者: 1. 《The Imperfective Paradox in Large Language Models》:通过“未完成体悖论”这一经典语言学现象测试大模型,发现开源模型普遍存在“目的论偏见”,即倾向于认为有目标的动作都会完成,揭示了模型更像“叙事预测引擎”而非逻辑推理者。 2. 《Memory efficiency and resource-rational encoding in sentence processing》:通过为Transformer模型添加工作记忆约束(注入噪声),迫使模型学习更高效地分配有限的记忆资源。结果发现,这种约束让模型的阅读节奏更接近人类,并促成了更压缩、更范畴化的表征。 3. 《Characterizing the Expressivity of Local Attention in Transformers》:利用形式语言理论解释了为何“局部注意力”机制在节省计算成本的同时,常能提升模型效果。研究表明,局部注意力引入了新的表达能力,与全局注意力互补,结合使用能获得最丰富的表达力。 此外,会议还评选了18篇杰出论文,华人在其中占据显著比例,尤其在强化学习、大模型安全与智能体等热门方向成果突出。 本届ACL数据显示,中国大陆作者占比高达54.0%,美国以18.4%位列第二。会议投稿量、审稿规模及参会人数均大幅增长,体现了该领域的空前繁荣与激烈竞争。

marsbit2小时前

ACL 2026华人霸榜,最佳论文一作全华人,杰出论文几乎包场

marsbit2小时前

交易

现货
活动图片