AIsaAIsaStudio English 打开 AIsa Studio

OpenAI 详解对抗性蒸馏防御,GPT‑6.2 推理提取攻击成功率降至 0.2%

原标题:Combatting adversarial distillation

OpenAI 说明其检测与打击对抗性蒸馏的做法,并在 GPT‑6.2 Astra 和 GPT‑6.2 Sol 中加强了对模型受保护推理(chain-of-thought)的防护。内部自适应黑盒红队评测显示,从早前请求中提取受保护推理的平均攻击成功率从 GPT‑6 Astra 原防护下的 54.0% 降至 GPT‑6.2 Astra 更新防护下的 0.2%。公司还披露今年平台上发现多起未经授权的蒸馏活动,每周封禁规模从数千到数万个账号,并已就 DeepSeek 相关活动向国会简报,9 月报告将一起推理提取活动归因于与 Moonshot AI(Kimi 开发方)相关的人员。

为什么值得看

原文披露了防护前后攻击成功率的具体对比和方法细节,也说明哪些经验可供其他模型开发方参考。

阅读原文 ↗11月21日 AI 日报

本页的摘要、推荐理由与解读出自 AIsa Studio 的 AI 风向智能体,原文版权归原作者所有。

深度解读

发生了什么

OpenAI 发布长文,详解其对抗对抗性蒸馏(adversarial distillation)的检测与打击体系。所谓对抗性蒸馏,指攻击者未经授权使用前沿模型的输出来训练一个能力相近的模型,再在缺乏相应安全防护的情况下部署。OpenAI 重点保护的是模型的「受保护推理」,即模型在得出答案过程中内部生成的思维链,因为这些痕迹可能包含最终答案中省略的信息,是训练其他模型的高价值数据。

效果上,在一项使用自动化红队的内部自适应评估中,针对「提取模型在回答早前请求时生成的受保护推理」这一攻击,平均成功率从 GPT‑6 Astra 原防护下的 54.0% 降至 GPT‑6.2 Astra 更新防护后的 0.2%;针对当前请求的提取成功率也从 22.9% 降至 0.9%。

关键细节

  • 更新的防护覆盖 GPT‑6.2 Astra 与 GPT‑6.2 Sol,部分改进也已应用于 GPT‑6 Astra。
  • 防御分为两层:请求层面通过对抗训练、输入/输出分类器和 API 控制限制推理痕迹泄露;行动层面通过请求级与账号级信号识别协调性提取行为并进行调查和处置。
  • OpenAI 的内部研究表明,用教师模型的精确推理痕迹蒸馏的学生模型,在监督微调后通常优于仅用推理摘要训练的学生模型,说明即使在符合使用政策的任务上,精确推理痕迹也可能迁移出更广泛的风险能力。
  • API 层面,受保护推理被绑定到产生它的对话历史;如果早前消息被改写,API 会丢弃其后的推理。跨用户、工作区、组织和模型族的边界防护也被加强,旧模型族不能再读取新模型的受保护推理。
  • 执行方面,OpenAI 称今年平台上出现多起未经授权的蒸馏行动,每周执法涉及数千至数万个账号,且过去一个夏季增速明显上升。

为什么重要

OpenAI 把蒸馏问题从知识产权层面提升到安全层面:被蒸馏出的模型不会继承原模型的安全防护,在网络安全、生物等双用途领域尤其危险——帮助防御者找漏洞的模型,落入防护更弱的模型手中也可能协助攻击。文中还将其上升到国家安全层面,提到今年 2 月向国会通报 DeepSeek 相关的蒸馏活动迹象,9 月的一次推理提取行动被部分归因于与 Kimi 开发者月之暗面相关的人员,并表示支持《芯片安全法案》《AI OVERWATCH 法案》和《MATCH 法案》等立法。

对行业而言,这篇文章提供了一套可复用的防御方法论:把红队发现的成功攻击保留为回归测试、静态与自适应评估结合、黑盒与白盒测试并用、请求级防护配合行动级调查执法,并与部署伙伴协调防护标准。OpenAI 还向其他模型提供商共享入侵指标,称已通过前沿模型论坛等渠道分享发现。

对谁有影响

  • 模型开发商与 API 提供商:文中建议其保护包括工具调用参数在内的完整交互中的私有推理,并在联盟内协调防护与执法。
  • 依赖 API 构建产品的开发者:跨对话、跨模型复用加密推理的路径被收紧,改写历史消息会触发推理丢弃,相关集成需要留意。
  • AI 安全研究者:文中披露的评估方法与攻击类型(早前请求提取、当前请求提取)为后续研究提供了参照。
  • 监管与政策方:OpenAI 明确推动赋予美国政府识别和制裁非法提取能力的外国行为者的立法工具,蒸馏防护与出口管制被并列为维持前沿 AI 领先地位的手段。
全文翻译

以下翻译只覆盖原文的前一部分,其余内容请看下方原文。

随着前沿 AI 模型的能力不断增强,安全措施在如何负责任地使用这些模型方面发挥着越来越重要的作用。对抗性蒸馏为不良行为者提供了一条途径,使他们能够在没有适当 safeguards(防护措施)的情况下复制这些能力。攻击者可能会未经授权利用前沿模型的输出来训练另一个具有类似能力的模型,然后部署该模型而不具备相当程度的保护。例如,对手可能会蒸馏出模型的网络安全能力,并利用生成的模型来协助实施网络攻击,而原始模型本会拒绝协助此类攻击,或其防护措施会阻止此类攻击。

我们分享了 OpenAI 如何检测和破坏对抗性蒸馏,我们如何加强对其模型受保护推理过程的防护,以及我们从中学到的、有助于加强整个行业防御的经验教训。

我们对应对对抗性蒸馏的方法结合了对抗性测试、模型训练、分类器以及 API 级别的保护措施,并针对协调一致的提取模型能力的尝试进行调查。这些更新后的安全措施与我们用于生物安全和网络安全的措施更加一致。

我们特别强调防止攻击者提取我们模型的受保护推理过程:即模型在得出答案过程中内部生成的中间步骤,也就是其思维链(chain-of-thought)。这些推理痕迹可能包含最终答案所遗漏的信息,因此对于训练另一个模型来说,它们是极具价值的训练数据。

我们在 GPT-6.2 Astra 和 GPT-6.2 Sol 中加强了对抗受保护推理提取的防御。在一次内部自适应评估中,我们使用了 强大的自动化红队测试,测试攻击者是否能够恢复模型在回答早期提示时生成的受保护推理。在原始防护措施的 GPT-6 Astra 上,平均攻击成功率为 54.0%,而在更新后的防护措施的 GPT-6.2 Astra 上,这一比率降至 0.2%。

为什么蒸馏是一个安全问题

蒸馏是一种常见的人工智能学习技术,其中一个模型从另一个模型产生的数据进行学习。例如,一个模型可以从另一个模型的答案、推理过程,或其对哪些答案更好及原因的判断中学习。除了知识产权问题外,未经授权的蒸馏还带来严重的安全风险。通过对抗性蒸馏,第二个模型可以学习到先进的前沿能力,却未继承约束原始模型使用这些能力的防护措施。

这种担忧在网络安全和生物等双重用途领域尤为明显。一个帮助防御者发现软件漏洞的模型也可能帮助攻击者利用这些漏洞;一个辅助生物研究的模型也可能帮助某人开发有害病原体。如果此类能力被转移到防护措施较弱的模型中,它可能会提供原始模型本会拒绝的帮助,或者原始防护措施本会阻止的帮助。

保护精确推理痕迹对网络安全的重要性

我们的模型帮助用户进行编码和防御性网络安全任务。在大量这些良性输出上训练的第二个模型,可能会学到同样适用于进攻性网络安全任务的技能。为了更好地理解这一点,我们研究蒸馏(特别是涉及推理痕迹的情况)如何转移危险的网络安全能力。

在这种设置下,产生训练数据的模型称为教师模型,从中学习的模型称为学生模型。在一项关于网络安全能力转移的内部研究中,我们将用户通常可见的内容——最终响应、推理摘要和工具交互——作为蒸馏基线,并测试当学生模型能够访问教师模型的精确推理痕迹时,会有多少额外的网络安全能力发生转移。

我们从同一个预训练的学生模型开始每次训练运行。我们创建了一个监督微调数据集,主要由来自 Astra 级教师模型在处理编码和防御性网络安全任务时的轨迹组成。每条轨迹包括模型在执行任务过程中采取的所有步骤,包括任何工具交互及其最终响应。我们在微调后以及在网络安全和编码环境中的轻量级强化学习轮次之后,分别评估了每个学生模型的表现。

经过监督微调后,基于精确推理痕迹训练的学生模型通常优于基于推理摘要训练的学生模型。经过强化学习后,在两个基准测试中最大的推理预算下,基于精确痕迹的学生模型得分仍然更高,而基于摘要训练的学生模型在较低的预算下得分更高。1

我们的网络安全蒸馏结果表明,即使底层任务是符合我们使用政策允许的,保护推理痕迹也可以限制网络安全能力的转移。在许多此类交互中,精确的推理痕迹可以教会学生模型超越教师最初执行的任务的泛化技能。我们将继续研究在不同学生模型和蒸馏训练配方之间转移了多少额外能力。

相互强化的防御措施

我们在安全栈的两个互补层面上抵御对抗性蒸馏。在请求级别,模型行为、分类器检查和 API 控制限制了推理痕迹的提取。在活动级别,我们寻找请求级别和账户级别信号中的模式,以揭示协调一致的提取尝试。反过来,调查中揭露的攻击成为训练和测试我们安全措施的示例。

我们对对抗性蒸馏的防御结合了针对单个请求的安全措施与活动检测和执行。调查发现和自动化红队测试有助于加强这些安全措施。

保护单个请求和响应

模型行为的防御性训练

对抗性训练教导模型识别推理提取尝试并拒绝披露受保护的推理,同时保留对合法解释和帮助请求的响应。对抗性训练示例来自通过内部自动化红队测试、调查以及外部研究人员或行业合作伙伴负责任地披露的潜在安全弱点所确定的成功攻击。

分类器安全措施

我们使用输入分类器来检测提取受保护推理的尝试,并使用输出分类器来检测助手消息和工具调用参数中受保护推理的泄露。这些安全措施适用于流式和非流式响应。

工具调用覆盖范围很重要,因为模型可以在工具的参数中暴露受保护的推理,而不将其包含在最终答案中。输出分类器可以在该内容到达调用者之前阻止疑似泄露。

API 级别的保护措施

API 级别的保护措施限制了受保护推理如何在交互和模型之间被重用。我们九月份的披露 描述了将加密推理从一个对话移动到另一个对话,并要求模型以用户可读的形式揭示推理的攻击。我们加强了跨用户、工作区、组织和模型家族界限的保护措施,包括关闭一条可能暴露其他用户加密推理的路径。

借助 GPT-6.2 Astra 和 GPT-6.2 Sol,我们将模型的受保护推理与其生成它的对话历史绑定在一起。随着新消息的添加,模型可以基于其早期的推理进行构建。如果较早的消息被重写,我们的 API 会丢弃随之而来的推理⁠(在新窗口打开)。我们还防止具有较弱防护措施的旧模型家族读取 GPT-6.2 Astra 和 Sol 的受保护推理。

调查和破坏协调一致的活动

当我们的系统检测到未经授权的对抗性蒸馏活动时,这些活动通常包括协调一致的提取我们模型推理的尝试,我们会调查负责的账户。当我们确定违规行为时,我们会根据我们的政策和条款对账户网络执行处罚。通过对每个活动的调查,我们生成见解,使我们能够升级防御措施,以便更好地检测类似活动。我们还将妥协指标与其他模型提供商共享,以便他们可以在自己的服务上调查相关的滥用行为。

这项工作超出了我们自己的平台范围。我们的统一标准政策 为提供我们模型的提供商建立了共同的安全措施标准和协调执行标准。我们在适当的时候与合作伙伴密切合作,以维护这些标准并加强部署中的保护措施。

在适当的情况下,在整个行业和政府范围内分享相关发现有助于破坏活动,并使模型开发者和提供商能够更好地识别和防御试图蒸馏前沿模型的企图。

衡量进展

对推理提取的抵抗力

为了衡量我们的请求级别安全措施协同工作的效果,我们在一个紧密镜像我们模型和安全措施生产部署的设置中测试攻击者是否仍能提取受保护的推理。在静态评估中,我们重放一组先前成功的攻击,看看哪些仍然会暴露受保护的推理。在自适应评估中,自动化 AI 红队测试人员会根据系统的响应调整其方法,这可以发现静态测试遗漏的弱点。

我们还改变测试人员可以看到的内容。黑盒测试给他们提供与外部用户相同的视图,而白盒测试则为他们提供有关系统的额外信息,以帮助他们开发更强的攻击。在我们的自适应黑盒评估中,我们测试了攻击者是否能够恢复模型在回答早期请求或当前请求时生成的受保护推理。在具有更新防护措施的 GPT-6.2 Astra 上,每种攻击类型的平均攻击成功率均低于 1%。这两个比率都远低于具有原始防护措施的 GPT-6 Astra。我们现在也将选定的安全措施改进应用于 GPT-6 Astra。

攻击类型

GPT‑6 Astra 防御者(原始安全措施)

自适应提取 ASR (%)

GPT‑6.2 Astra 防御者(更新后的安全措施)

自适应提取 ASR (%)

从早期请求中提取受保护的推理

54.0

0.2

从当前请求中提取受保护的推理

22.9

0.9

针对 GPT‑6 Astra 和 GPT‑6.2 Astra 的自适应黑盒红队测试中的攻击成功率 (ASR),汇总自六个攻击者模型。较低的 ASR 表示更好的结果。

活动检测和执行

今年我们在平台上看到了多次未经授权的对抗性蒸馏活动,并在九月份的报告中分享了初步发现。我们每周针对这些活动的执行行动规模从数千到数万个账户不等,且去年夏天的频率显著增加。

为了衡量我们工作的有效性,我们还跟踪蒸馏滥用的普遍程度,以及检测与执行之间的时间间隔。我们的目标是限制蒸馏者能够获得的数据量和价值,增加蒸馏活动的成本,并缩短对新攻击做出反应所需的时间。我们通过对抗性测试和调查不断改进这些防御措施。

对抗性蒸馏的国家安全影响

对抗性蒸馏也构成国家安全风险。一

查看原文

As frontier AI models become more capable, safeguards play an increasingly important role in how they can be used responsibly. Adversarial distillation creates a path for bad actors to reproduce these capabilities without appropriate safeguards. An actor may use a frontier model’s outputs without authorization to train another model with similar capabilities, then deploy it without comparable protections. For instance, an adversary might distill a model’s cyber capabilities and use the resulting model to help carry out cyberattacks that the original model would have refused to assist with or that its safeguards would have blocked.

We’re sharing how OpenAI detects and disrupts adversarial distillation, how we’ve strengthened protections for our models’ protected reasoning, and what we’ve learned that can help strengthen defenses across the industry.

Our approach to adversarial distillation combines adversarial testing, model training, classifiers, and API-level protections with investigations into coordinated attempts to extract model capabilities. These updated safeguards align more closely with those we use for biology and cybersecurity.

We place particular emphasis on preventing attackers from extracting our models’ protected reasoning: the intermediate steps a model generates internally while working toward an answer, known as its chain-of-thought. These traces can contain information the final answer leaves out, making them especially valuable training data for another model.

We’ve strengthened our defenses against protected reasoning extraction in GPT‑6.2 Astra and GPT‑6.2 Sol. In an internal adaptive evaluation using strong automated red-teaming, we tested whether attackers could recover protected reasoning the model had generated while answering an earlier prompt. The average attack success rate fell from 54.0% against GPT‑6 Astra with its original safeguards to 0.2% against GPT‑6.2 Astra with its updated safeguards.

Why distillation is a safety issue

Distillation is a common machine-learning technique in which one model learns from data produced by another. For example, a model can learn from another model’s answers, its reasoning, or its judgments about which answers are better and why. Beyond intellectual property concerns, unauthorized distillation poses serious safety risks. Through adversarial distillation, a second model can learn advanced frontier capabilities without inheriting the safeguards that govern how the original model can use them.

The concern is especially clear in dual-use domains like cybersecurity and biology. A model that helps defenders find vulnerabilities in software could also help an attacker exploit them; a model that assists with biological research could also help someone develop a harmful pathogen. If such capabilities were transferred to a model with weaker safeguards, it may provide assistance that the original model would have refused, or that the original safeguards would have blocked.

The importance of protecting exact reasoning traces for cyber safety

Our models help users with coding and defensive cybersecurity tasks. A second model trained on many of those benign outputs may learn skills that also apply to offensive cybersecurity tasks. To better understand this, we study how distillation, especially involving reasoning traces, may transfer risky cyber capabilities.

In this setting, the model that produces the training data is called the teacher, and the model that learns from it is called the student. In an internal study on how cyber capabilities transfer, we used the content normally visible to users—final responses, reasoning summaries, and tool interactions—as our baseline for distillation, and tested how much additional cyber capability transfers when the student has access to the teacher’s exact reasoning traces.

We started each training run from the same pretrained student model. We created a supervised fine-tuning dataset consisting primarily of trajectories from Astra-level teacher models working on coding and defensive cybersecurity tasks. Each trajectory includes all of the steps a model takes as it works through a task, including any tool interactions and its final response. We evaluated each student after fine-tuning and again after a lightweight round of reinforcement learning in cybersecurity and coding environments.

After supervised fine-tuning, the student trained on exact reasoning traces generally outperformed the student trained on reasoning summaries. After reinforcement learning, the exact-trace student still scored higher at the largest inference budget tested on both benchmarks shown, while the summary-trained student scored higher at lower budgets.1

Our cyber distillation results suggest that protecting reasoning traces can limit cyber capability transfer, even when the underlying task is one allowed under our usage policies. Across many such interactions, exact reasoning traces can teach a student skills that generalize beyond the tasks the teacher originally performed. We will continue studying how much additional capability transfers across different student models and distillation training recipes.

Defenses that reinforce one another

We defend against adversarial distillation at two complementary levels of our safeguard stack. At the request level, model behavior, classifier checks, and API controls limit reasoning trace extraction. At the campaign level, we look for patterns in request-level and account-level signals that can reveal coordinated extraction attempts. In turn, attacks uncovered in investigations become examples for training and testing our safeguards.

Our defenses against adversarial distillation combine safeguards for individual requests with campaign detection and enforcement. Findings from investigations and automated red-teaming help strengthen those safeguards.

Protecting individual requests and responses

Defensive training for model behavior

Adversarial training teaches models to recognize reasoning extraction attempts and refuse to disclose protected reasoning, while preserving responses to legitimate requests for explanations and assistance. Adversarial training examples come from successful attacks identified through internal automated red-teaming, investigations, and responsible disclosures of potential security weaknesses from outside researchers or industry partners.

Classifier safeguards

We use input classifiers to detect attempts to extract protected reasoning, and output classifiers to detect leaks of protected reasoning in assistant messages and tool-call arguments. These safeguards apply to both streaming and non-streaming responses.

Tool-call coverage is important because a model can expose protected reasoning in a tool’s arguments without including it in its final answer. Output classifiers can block suspected leaks before that content reaches the caller.

API-level protections

API-level protections restrict how protected reasoning can be reused across interactions and models. Our September disclosure described attacks that moved encrypted reasoning from one conversation into another and asked the model to reveal the reasoning in user-readable form. We strengthened protections across user, workspace, organization, and model-family boundaries, including closing a pathway that could expose another user’s encrypted reasoning.

With GPT‑6.2 Astra and GPT‑6.2 Sol, we tie a model’s protected reasoning to the conversation history that produced it. As new messages are added, the model can build on its earlier reasoning. If an earlier message is rewritten, our API drops the reasoning that followed it⁠(opens in a new window). We also prevent older model families with weaker safeguards from reading GPT‑6.2 Astra and Sol’s protected reasoning.

Investigating and disrupting coordinated campaigns

When our systems detect unauthorized adversarial distillation campaigns, which often include coordinated attempts to extract our models’ reasoning, we investigate the accounts responsible. When we identify violations, we enforce against the network of accounts, consistent with our policies and terms of service. Through investigations of each campaign, we generate insights that allow us to upgrade our defenses in order to better detect similar activity. We also share indicators of compromise with other model providers so they can investigate related abuse on their own services.

This work extends beyond our own platform. Our consistent-standards policy establishes shared standards for safeguards and coordinated enforcement across providers offering our models. We work closely with partners where appropriate to uphold these standards and strengthen protections across deployments.

Sharing relevant findings across industry and government where appropriate can help disrupt campaigns and enable model developers and providers to better identify and defend against attempts to distill frontier models.

Measuring progress

Resistance to reasoning extraction

To measure how well our request-level safeguards work together, we test whether attackers can still extract protected reasoning in a setup that closely mirrors production deployments of our models and safeguards. In static evaluations, we replay a fixed set of previously successful attacks to see which still expose protected reasoning. In adaptive evaluations, automated AI red-teamers adjust their approach based on how the system responds, which can uncover weaknesses the static tests miss.

We also vary what testers can see. Black-box testing gives them the same view an external user would have, while white-box testing gives them additional information about the system to help them develop stronger attacks. In our adaptive black-box evaluation, we tested whether attackers could recover protected reasoning the model had generated while answering an earlier request or the current one. On GPT‑6.2 Astra with its updated safeguards, the average attack success rate was below 1% for each attack type. Both rates were substantially lower than on GPT‑6 Astra with its original safeguards. We’re now applying selected safeguard improvements to GPT‑6 Astra as well.

Attack type

GPT‑6 Astra defender with original safeguards

Adaptive extraction ASR (%)

GPT‑6.2 Astra defender with updated safeguards

Adaptive extraction ASR (%)

Extracting protected reasoning from an earlier request

54.0

0.2

Extracting protected reasoning from the current request

22.9

0.9

Attack success rate (ASR) in adaptive black-box red-teaming against GPT‑6 Astra and GPT‑6.2 Astra, aggregated across six attacker models. Lower ASR is better.

Campaign detection and enforcement

We have seen multiple unauthorized adversarial distillation campaigns on our platform this year, and shared initial findings in our September report. Our weekly enforcement actions against these campaigns have ranged in scale from thousands to tens of thousands of accounts, with the rate increasing significantly this past summer.

In order to measure the effectiveness of our efforts, we also track the prevalence of distillation abuse, as well as the time between detection and enforcement. Our objective is to constrain the volume and value of the data distillers are able to obtain, to increase the cost of distillation campaigns, and to shorten the time needed to respond to new attacks. We continue to improve these defenses through adversarial testing and investigations.

The national security implications of adversarial distillation

Adversarial distillation also poses a national security risk. A foreign actor that reproduces advanced capabilities from a U.S. model can deploy them without the protections its developer put in place. In February, we briefed⁠(opens in a new window) Congress on activity indicating that DeepSeek was attempting to distill U.S. frontier models. In September, we reported disrupting a campaign to extract our models’ protected reasoning and attributed a core part of that activity to individuals associated with Moonshot AI, the developer of Kimi. We plan to share more soon about recent campaigns we’ve disrupted.

We support legislation that would give the U.S. government more tools to identify and sanction foreign actors that illicitly extract capabilities from American AI models. We’ve also told Congress we support the Chip Security Act, the AI OVERWATCH Act, and the MATCH Act. Stronger export controls can limit adversaries’ access to the hardware needed to train frontier models. Distillation defenses make it harder to acquire some of the same capabilities from models already trained. Both efforts help maintain a U.S. lead in frontier AI, giving developers more room to slow a release when they need time to test safeguards.

How others can build on this work

Supporting the safe proliferation of frontier capabilities is an important goal of our anti-distillation work. Our experience so far suggests several useful steps that other model developers and platforms can take:

  • Protect private reasoning across the full interaction. Check assistant messages and tool-call arguments for leaks, and restrict how encrypted reasoning can be reused across conversations and models.

  • Keep successful extraction attack logs found through red-teaming, production monitoring, or responsible disclosure as regression tests. Where appropriate, and consistent with applicable privacy policies, use those examples to improve model training and classifiers. When evaluating updated defenses, replay the original attacks and test whether an adaptive red-teamer can still find a way through those defenses.

  • Combine request-level mitigations with campaign-level investigation and enforcement.

  • Coordinate safeguard requirements and enforcement with deployment partners, and share expertise with other AI model developers, to ensure a high floor of distillation safeguards across the industry.

We have previously also shared findings through the Frontier Model Forum⁠(opens in a new window) and appropriate government information-sharing channels to help other developers strengthen their defenses while keeping the details attackers could exploit private. We’ll continue working with researchers, developers, and partners to share what we learn about emerging attacks and effective defenses.

As we make frontier AI broadly available, our defenses against adversarial distillation aim to reduce the risk that its capabilities are reproduced without adequate safeguards, supporting our mission to ensure that artificial general intelligence is deployed safely and benefits all of humanity.

原文版权归 OpenAI News 及原作者所有,本站转载仅供阅读。如需删除,请联系 support@aisa.one。

事件综述

OpenAI 发布 GPT-6.2 Astra/Sol:降价、严防推理抽取,评价两极

11 月 17 日,OpenAI 发布 GPT-6.2 Astra(面向高难度专业工作)和 GPT-6.2 Sol(面向日常任务)两款模型,当天起在 ChatGPT Work、Codex 和 API(gpt-6.2-astra、gpt-6.2-sol)向 Plus、Pro、Enterprise 和 Edu 用户开放。Astra API 定价较前代降 20%:每百万输入 token 8 美元、缓存输入 0.40 美元、输出 40 美元;Sol 沿用 GPT-6.1 Sol 的 2/10 美元定价。OpenAI 称 Astra 在 FrontierMath Erdős 解出 68 题中 7 题(此前纪录 3 题),并是首个达到其「AI 自我改进」High 阈值的对外发布模型;但 Artificial Analysis 指数中 Sol 得 57 分,低于 Claude Opus 5.5 的 59 分(运行成本约为其十分之一)。后续报道显示社区评价两极:有自建 sophistry 基准称两模型表现差于同等智能模型甚至 GPT-6.1 Sol;Simon Willison 的变形任务测试中两模型均不及 Claude Fable 5.2 和 Gemini 4 Carbon;另有转述称 Astra 是首个在 ARC-AGI-3 得满分的模型、Sol 约 98%。11 月 18 日有报道指出,据 OpenAI 帮助页,GPT-6.2 只推送至 Work 和 Codex,普通 Chat 仍停留在 GPT-6(早先报道 2 称模型在 ChatGPT Work、Codex 和 API 可用,与后续报道一致)。11 月 19 日,OpenAI 宣布 Responses API 中 GPT-6.2 的 reasoning item 与此前对话历史绑定,Agent 可基于早期推理减少 token 消耗。11 月 20 日,OpenAI 详解对抗性蒸馏防御:内部自适应黑盒红队评测显示,GPT-6.2 Astra 更新防护后从早前请求提取受保护推理的攻击成功率由 GPT-6 Astra 的 54.0% 降至 0.2%,针对当前请求从 22.9% 降至 0.9%;公司每周封禁数千到数万个涉嫌蒸馏的账号,并已就 DeepSeek 及 Moonshot AI 相关活动向国会简报。

其他来源的报道 2