他认为，模型通过用户的提示词、智能体工具使用及纠错反馈进行学习，这些"废气"数据提炼出的机构知识是竞争对手无法买到的。 纳德拉批评模型厂商一方面自由抓取互联网数据训练，另一方面却限制企业对模型进行"知识蒸馏"。 他建议企业保留对提示词、反馈等数据的所有权，在云端构建专有学习环境，并采用"编排层"工具在不同AI模型间灵活切换，避免被单一厂商锁定。
在所有关于人工智能潜在缺点的激烈辩论中，有一个担忧在硅谷的人工智能爱好者中引起了最多的焦虑。他们担心的是，那些出售专有模型的巨型人工智能实验室在某种程度上就像是特洛伊木马一样。
这种担忧在于，当初创公司和企业使用来自OpenAI和Anthropic等实验室的人工智能模型时，这些实验室会获得越来越多的这些公司最敏感的商业信息。模型制造商随后可以利用这些知识为自己谋利，可能成为自己客户的竞争对手。发出这种警告的人从风投人士如Jason Calacanis（https://techcrunch.com/2026/05/20/sam-altman-makes-mic-drop-offer-to-every-y-combinator-startup/?utm_source=chatgpt.com）到Palantir首席执行官Alex Karp都有。
现在，在一篇令人惊讶的博客文章中（https://snscratchpad.com/posts/reverse-information-paradox/，发布于周日），微软首席执行官萨蒂亚·纳德拉加入了这个行列。纳德拉警告称，人工智能用户（他称之为“买家”）正在支付双重代价。他们心甘情愿地为人工智能令牌使用支付费用，但同时也在不自觉中交出宝贵的数据。
“你实际上为智能付出两次代价，一次是用钱支付，另一次是用更有价值的东西支付：为了让智能有用而必须透露的专有知识。你希望模型表现得越好，就要提供越多这样的知识！”他写道。
他认为，最危险的是，企业实际上正在教模型了解他们业务的细微差别。
“模型从‘排放物’中学习，人们写的提示、代理使用的工具，尤其是在模型出错时人们进行的纠正。每一次纠正都会被提炼为机构性的专业知识，”他写道。
这是“竞争对手永远买不到的那种知识”，然而企业却在把它交出去。
纳德拉认为，如果人工智能公司可以自由地抓取互联网来训练他们的模型，那么企业有权研究——或“蒸馏”——这些模型也是公平的。“蒸馏”是指使用模型自身的输出了解其工作原理，并基于这些见解训练一个新的、通常更便宜的模型。二月份，Anthropic 指控中国开源模型向 Claude 发送了数百万个提示：https://techcrunch.com/2026/02/23/anthropic-accuses-chinese-ai-labs-of-mining-claude-as-us-debates-ai-chip-exports/，以此来改进它们自己的模型，并敦促美国政府加强出口管制。
纳德拉的观点是，模型制造商不能鱼与熊掌兼得。他们自由地用全世界的数据来训练自己模型，却限制别人对他们的模型做同样的事，这很虚伪。
“虽然模型提供者拥有在公共数据上训练模型的合理使用权带来的巨大创新是必要的，但我觉得讽刺的是，现状却是反过来对蒸馏施加限制性条款，”纳德拉写道。
当模型制造商“保留从客户使用和交互数据中学习的权利”时，纳德拉尤其关心。
纳德拉的解决方案是巨型云服务提供商的首席执行官可能会提出的那种建议。他希望公司“保留其数据的所有权”，包括提示、反馈等。因此，他敦促它们在云端（数据很可能已经存储在这里，而且方便地可能是微软的云 Azure）构建自己的“专有学习环境”。他还希望公司建立他所谓的“编排层”——本质上是一种可以轻松切换不同供应商的 AI 模型的方法，而不是被锁定在单一模型中。允许公司做到这一点的 AI “网关”等工具正变得越来越流行。
虽然纳德拉从未使用“开源”一词来表示保留所有权的方法，但这是一个显而易见的潜台词。然而，还有另一个潜台词。
许多大型公司除了使用云服务外，仍拥有自己的数据中心，它们已经开始向安装在本地的开源模型（业内术语称为“on-prem”）迁移。Solo.io 的创始人兼首席执行官 Idit Levine——该公司开发网络和安全软件，帮助企业管理 AI 系统——表示，她正亲眼看到这一变化在她的客户中发生。在尝试了专有模型供应商之后，客户会开始问自己：“我能否使用开源模型并在本地运行？它几乎可以完成大型模型 90% 的功能，而且成本会低得多。”她告诉 TechCrunch：“他们理解这一点，并且可以控制它。”
去年，Solo.io 的技术被选为 Linux 基金会 Agentgateway 项目的技术支持：https://www.linuxfoundation.org/press/linux-foundation-welcomes-agentgateway-project-to-accelerate-ai-agent-adoption-while-maintaining-security-observability-and-governance。她的公司拥有 T-Mobile、ADP 和 SAP 等企业客户。她看到企业越来越多地安装本地开源模型，并认为这是企业 AI 应用的下一大浪潮。
她并非孤例。Vercel（最著名的是一个用于构建和托管网站的平台，最近增加了 AI 模型切换工具）和 OpenRouter（一个帮助开发者在不同 AI 模型间路由请求的公司）都在看到开源模型流量的激增：https://techcrunch.com/2026/07/07/why-the-rise-of-open-source-ai-isnt-hurting-anthropic-yet/。事实上，上个月开源模型占 Vercel 网关路由的所有流量的 29%：https://x.com/vercel/status/2076712644539543821。
随着微软 CEO——这家公司在 OpenAI 和 Anthropic 都有投资——现在公开建议企业谨慎使用专有模型，我们可以预计这一趋势将继续增长。Nadella 写道：“在消耗智能时，你也在创造智能。而你所创造的，应当属于你自己。”
当你通过我们文章中的链接进行购买时，我们可能会获得一小部分佣金：https://techcrunch.com/techcrunch-affiliate-monetization-standards/。这不会影响我们的编辑独立性。
TechCrunch创始人峰会最后机会，享受最高190美元的优惠。加入1000名各阶段的创始人和风险投资人，获得实际的扩展见解和能推动成果的联系。优惠截止至6月26日晚上11:59（太平洋时间）。
德克萨斯州致命车祸的特斯拉司机踩下油门100%，NTSB证实：https://techcrunch.com/2026/07/15/tesla-driver-in-fatal-texas-crash-pressed-accelerator-100-ntsb-confirms/ Sean O'Kane：https://techcrunch.com/author/sean-okane/
在硬件法律争端中，OpenAI发布售价230美元的Codex键盘：https://techcrunch.com/2026/07/15/amid-hardware-legal-battle-openai-releases-a-230-keyboard-for-codex/ Lucas Ropek：https://techcrunch.com/author/lucas-ropek/
OpenAI首款硬件设备据报道是一个无屏幕可移动的扬声器：https://techcrunch.com/2026/07/14/openais-first-hardware-device-is-reportedly-a-screenless-speaker-that-can-move/ Lucas Ropek：https://techcrunch.com/author/lucas-ropek/
Anthropic最新广告让人毛骨悚然：https://techcrunch.com/2026/07/14/anthropics-newest-ad-is-creeping-people-out/ Lucas Ropek：https://techcrunch.com/author/lucas-ropek/
Satya Nadella向使用AI的公司发出令人震惊的警告：https://techcrunch.com/2026/07/13/satya-nadella-has-issued-a-shocking-warning-to-companies-using-ai/ Julie Bort：https://techcrunch.com/author/julie-bort/
苹果针对OpenAI的商业秘密诉讼中的最疯狂指控：https://techcrunch.com/2026/07/13/the-wildest-allegations-in-apples-trade-secrets-lawsuit-against-openai/ Sarah Perez：https://techcrunch.com/author/sarah-perez/
Anthropic开始为印度本地化Claude定价，这也是除美国外最大的市场：https://techcrunch.com/2026/07/13/anthropic-starts-localizing-claude-pricing-for-india-its-biggest-market-after-the-us/ Jagmeet Singh：https://techcrunch.com/author/jagmeet-singh/
Of all the debates raging about the potential downsides of AI, there is one worry causing the most hand-wringing among AI enthusiasts in Silicon Valley. Their fear is that the giant AI labs that sell proprietary models are somehow acting like Trojan horses.
The concern is that, as startups and enterprises use AI models from labs like OpenAI and Anthropic, the labs gain ever-increasing access to those companies’ most sensitive business information. The model makers can then use that knowledge for themselves, potentially becoming competitors to their own customers. Those issuing such warnings range from VCs like Jason Calacanis：https://techcrunch.com/2026/05/20/sam-altman-makes-mic-drop-offer-to-every-y-combinator-startup/?utm_source=chatgpt.com to Palantir CEO Alex Karp.
Now, in a surprising blog post：https://snscratchpad.com/posts/reverse-information-paradox/ published on Sunday, Microsoft CEO Satya Nadella has joined this crowd. Nadella warns that AI users (the “buyers” as he calls them) are paying twice. They knowingly spend for AI token usage but they also, obliviously, hand over valuable data in the process.
“You essentially pay for intelligence twice, once with money, and again with something even more valuable: the proprietary knowledge you must reveal to make that intelligence useful. The better you want the model to perform, the more of that knowledge you have to feed it!” he writes.
Most dangerously, enterprises are literally teaching the models about the nuances of their businesses, he argues.
“Models learn from ‘exhaust,’ the prompts people write, the tools agents use, and especially the corrections people make when the model is wrong. Every correction is distilled into institutional know-how,” he writes.
This is “the kind of knowledge a competitor could never buy,” and yet enterprises are handing it over.
Nadella argues that if AI companies get to freely scrape the internet to train their models, it’s only fair that enterprises get to study — or “distill” — those models in return. “Distillation” is the practice of using a model’s own outputs to learn how it works and to train a new, often cheaper, model based on those insights. In February, Anthropic accused Chinese open source models of sending millions of prompts to Claude：https://techcrunch.com/2026/02/23/anthropic-accuses-chinese-ai-labs-of-mining-claude-as-us-debates-ai-chip-exports/ as a way to improve their own models, and urged the U.S. government crack down on export controls.
Nadella’s point is that model makers can’t have it both ways. It’s hypocritical for them to freely train on the world’s data while restricting others from doing the same to their models.
“While the great innovation that comes from model providers having fair use rights to train models on public data is needed, I find it ironic that the status quo is to then turn around and impose restrictive terms on distillation,” Nadella writes.
Nadella is particularly concerned when model makers “reserve the right to learn from customer usage and interaction data.”
Nadella’s solution is the kind of thing the CEO of a giant cloud provider would suggest. He wants companies to “retain ownership” of their data, including prompts, feedback, etc. So he’s urging them to build their own “proprietary learning environments” on the cloud (where their data is likely already stored anyway and, conveniently, could mean Microsoft’s cloud, Azure). He also wants companies to build in what he calls “orchestration layers” — essentially, a way to easily switch between AI models from different providers rather than being locked into one. Tools like AI “gateways” that let companies do exactly this have become increasingly popular.
While Nadella never uses the words “open source” as the method for retaining ownership, this is an obvious subtext. Yet, there’s another subtext.
Large companies, many of which still have some of their own data centers in addition to using the cloud, are already moving to open source models installed on their own premises (“on-prem,” in industry jargon). Idit Levine, founder and CEO of Solo.io — which makes networking and security software that helps enterprises manage AI systems — says she’s seeing exactly this shift play out with her own customers. After experimenting with proprietary model makers, they start asking themselves: “Can I take an open source model and run it on-prem? It will do almost 90% of what the big one’s doing. It will cost way less,” she tells TechCrunch. “They understand that, and they can control it.”
Solo.io’s technology was selected last year to be the tech powering the Linux Foundation’s Agentgateway project：https://www.linuxfoundation.org/press/linux-foundation-welcomes-agentgateway-project-to-accelerate-ai-agent-adoption-while-maintaining-security-observability-and-governance. Her company counts enterprises like T-Mobile, ADP, and SAP as customers. She sees companies increasingly installing on-premise open source models and sees it as the next big wave in enterprise AI use.
She’s not alone. Vercel (best known as a platform for building and hosting websites, which has recently added AI model-switching tools) and OpenRouter (a company that helps developers route requests across different AI models) are both seeing a surge in traffic to open source models：https://techcrunch.com/2026/07/07/why-the-rise-of-open-source-ai-isnt-hurting-anthropic-yet/. In fact, open models accounted for 29% of all traffic routed through Vercel’s gateway：https://x.com/vercel/status/2076712644539543821 last month.
With the CEO of Microsoft, a company that has invested in both OpenAI and Anthropic, now openly urging enterprises to be wary of using proprietary models, we’ll bet this trend continues to grow. “In consuming intelligence, you are creating intelligence. And what you create should belong to you,” Nadella writes.
When you purchase through links in our articles, we may earn a small commission：https://techcrunch.com/techcrunch-affiliate-monetization-standards/. This doesn’t affect our editorial independence.
Last chance to save up to $190 on TechCrunch Founder Summit. Join 1,000+ founders and VCs at all stages for real-world scaling insights and connections that move the needle. Savings end June 26, 11:59 p.m. PT .
Tesla driver in fatal Texas crash pressed accelerator 100%, NTSB confirms：https://techcrunch.com/2026/07/15/tesla-driver-in-fatal-texas-crash-pressed-accelerator-100-ntsb-confirms/ Sean O'Kane：https://techcrunch.com/author/sean-okane/
Amid hardware legal battle, OpenAI releases a $230 keyboard for Codex：https://techcrunch.com/2026/07/15/amid-hardware-legal-battle-openai-releases-a-230-keyboard-for-codex/ Lucas Ropek：https://techcrunch.com/author/lucas-ropek/
OpenAI’s first hardware device is reportedly a screenless speaker that can move：https://techcrunch.com/2026/07/14/openais-first-hardware-device-is-reportedly-a-screenless-speaker-that-can-move/ Lucas Ropek：https://techcrunch.com/author/lucas-ropek/
Anthropic’s newest ad is creeping people out：https://techcrunch.com/2026/07/14/anthropics-newest-ad-is-creeping-people-out/ Lucas Ropek：https://techcrunch.com/author/lucas-ropek/
Satya Nadella has issued a shocking warning to companies using AI：https://techcrunch.com/2026/07/13/satya-nadella-has-issued-a-shocking-warning-to-companies-using-ai/ Julie Bort：https://techcrunch.com/author/julie-bort/
The wildest allegations in Apple’s trade secrets lawsuit against OpenAI：https://techcrunch.com/2026/07/13/the-wildest-allegations-in-apples-trade-secrets-lawsuit-against-openai/ Sarah Perez：https://techcrunch.com/author/sarah-perez/
Anthropic starts localizing Claude pricing for India, its biggest market after the US：https://techcrunch.com/2026/07/13/anthropic-starts-localizing-claude-pricing-for-india-its-biggest-market-after-the-us/ Jagmeet Singh：https://techcrunch.com/author/jagmeet-singh/
扫码直达文章详情