11-19 更新 实时
首页/ 深度/ 正文

两天内发现28个零日漏洞,机器人安全跟得上吗

深度 来源:IEEE Spectrum AI 2026-11-16 18:00 0 阅读 机器翻译
两天内发现28个零日漏洞,机器人安全跟得上吗

AI 在物理机器人软件中发现 28 个零日漏洞 - IEEE Spectrum

不到两天发现 28 个零日漏洞。机器人安全跟得上吗?

献给技术内行

献给技术内行

IEEE Spectrum 是 IEEE 的旗舰出版物——IEEE 是全球最大的致力于工程与应用科学的专业组织。我们的文章、视频和信息图表让读者了解技术、工程和科学领域的最新进展。

创建账户,享受更多免费内容与权益

保存文章以便稍后阅读需要 IEEE Spectrum 账户

The Institute 的内容仅对会员开放

下载完整 PDF 期刊仅限 IEEE 会员

下载本电子书仅限 IEEE 会员

访问 Spectrum 数字版仅限 IEEE 会员

关注主题是 IEEE 会员专属功能

加入全球最大的致力于工程与应用科学的专业组织,即可访问 Spectrum 的全部文章、档案、PDF 下载及其他权益。进一步了解 IEEE →

加入全球最大的致力于工程与应用科学的专业组织,即可获取本电子书以及 IEEE Spectrum 的全部文章、档案、PDF 下载及其他权益。进一步了解 IEEE →

访问数千篇文章——完全免费

创建账户即可获取专属内容和功能:保存文章、下载合集、发表评论——全部免费!如需完整访问权限和权益,请订阅 Spectrum。

不到两天发现 28 个零日漏洞。机器人安全跟得上吗?

VicOne 的 AI 辅助研究揭示了更快的零日漏洞发现对机器人安全意味着什么

VicOne 是一家 Physical AI 网络安全公司,保护车辆和机器人免受可能改变其行为、危及安全的网络威胁。

VicOne 利用 AI 在 Physical AI 工具中发现 28 个零日漏洞——这些弱点可能影响机器人训练、测试和真实环境中的控制。

本文由 VicOne 提供。

VicOne 利用 AI 辅助研究,在不到两天的时间里,在支撑 Physical AI 生态的软件和工具中初步识别出 28 个零日发现。这些项目包括由 Google、NVIDIA 和 Meta 等开发或维护的项目。

基于初步验证,全部 28 个发现都被归类为零日漏洞。在厂商和漏洞管理机构审查后,数量可能会变化,且部分发现尚未获得 CVE 跟踪编号。

这些工具中有许多帮助团队训练、仿真、测试和操作机器人。其中的弱点可能影响团队构建的模型、他们信任的测试结果,以及发送给机器人的指令。

这些发现对机器人团队提出了两项要求:在 AI 帮助更快发现弱点的同时做出响应,以及评估每一个弱点可能如何影响机器人的安全与运行。

AI 给漏洞发现带来的变化

漏洞发现涉及检查代码、追踪系统之间的通信方式,以及测试某个疑似弱点是否真的可被利用。AI 可以加速其中部分工作,帮助研究人员在更短时间内分析更多软件。攻击者也可能利用类似的能力。

当弱点是零日漏洞——此前未知且尚无可用修复——时,更快的发现可能让团队几乎没有时间响应。虽然这种暴露并不必然意味着已经发生攻击,但团队仍需确定自己是否使用了受影响的软件,以及攻击者可能访问或更改什么。

这些发现处于机器人生命周期的哪些环节

VicOne 的发现横跨机器人生命周期的三个阶段:开发与训练、仿真与验证,以及部署与运行时。每个阶段的弱点都可能影响机器人学到什么、如何被测试,以及如何被控制。

图 1。在这组研究中,75% 的发现涉及开发、训练、仿真或验证工具,凸显了机器人部署之前就可能出现的风险。VicOne

以下场景展示了被攻陷的组件可能如何在各个阶段影响机器人。

开发与训练:隐藏的改动可能一路进入成品机器人。团队下载预训练模型,为自家机器人做适配。加载恶意文件可能让攻击者获得训练系统的访问权限,从而篡改训练数据或模型本身。被操纵的模型在常规测试中可能表现正常,但在部署后会对某个特定物体、场景或指令做出错误响应。风险可能经由团队信任的工具进入,而无需任何人直接攻击机器人。

仿真与验证:测试通过可能带来虚假的信心。导入被污染的仿真场景或机器人模型,可能让攻击者获得仿真环境的访问权限。若控制程度足够,攻击者可以篡改仿真传感器读数、物理条件或测试结果。这样一来,机器人可能看似通过了本应暴露其不安全行为的测试,使团队基于误导性证据批准其投入真实环境使用。

部署与运行时:操作人员可能失去可见性或控制权。攻击者可能利用弱点修改影响运动规划的文件,或发送恶意地图数据导致操作员的监控显示崩溃。远程操作计算机被攻陷,也可能使发送给机器人的指令面临风险。视系统及其防护机制而定,这些弱点可能改变机器人的运动,或让操作员无法看到正在发生的情况并及时响应。

这些都是潜在后果,并非每一项发现都已得到实证。实际影响取决于攻击者拥有何种访问权限、系统的设计,以及其他防护机制能否阻止这些改动传导到物理行为。

在另一项演示中,VicOne 的研究人员给出了一个影响防护功能的弱点的具体案例。研究人员利用一个漏洞,使机器人的远程停止指令失效。这一案例说明,当支撑防护机制的软件遭到破坏时,团队为何需要测试这些防护措施是否仍然有效。

视频 1。VicOne 研究人员演示了一个可使机器人远程停止指令失效的漏洞。VicOne

已发布的补丁仍需证明其安全

AI 可以缩短寻找弱点的时间,但机器人团队仍需判断某个修复方案是否可以安全部署。即便软件维护者发布了补丁,团队也必须找出受影响的系统、检查兼容性,并验证更新是否保持了预期行为。

此前一次成功的测试仍然是有用的证据。然而,新发现的弱点可能暴露出那次测试从未考察过的情况。

有三个问题可以帮助团队应对:

我们在哪些地方使用了受影响的软件?包括训练用计算机、仿真环境和已部署的机器人。

这个弱点可能改变什么?评估它是否会影响模型、测试结果、操作人员的可见性,或机器人的行为。

我们如何验证防护是否有效?确认修复措施解决了该弱点,并确认更新后的系统仍按预期运行。

当立即更新并不现实时,团队可能需要采取临时防护措施,例如限制对受影响服务的访问,同时准备经过验证的修复方案。

从发现漏洞到更安全的部署

随着这 28 项发现的详细信息通过协同披露逐步公开,机器人制造商接下来要做的,是把每一项发现与其在开发、部署和机器人行为方面可能产生的影响对应起来。

这项工作还要求理解威胁可能如何影响机器人行为,以及如何保护已在运行中的系统。VicOne Radeis 帮助团队识别潜在的零日漏洞、评估 AI 模型与系统安全,并通过仿真在部署前评估这些风险可能如何影响机器人行为。VicOne Rthena 则在运行期间支持在已批准策略范围内的威胁检测与响应。

图 2。网络安全在整个生命周期中强化机器人安全证据。VicOne 帮助团队在部署前测试安全限值抵御操纵的能力,并在运行中监测支撑安全行为的各项条件。VicOne

只有当团队能把发现转化为行动时,更快的漏洞发现才有意义:确定哪些需要关注、验证防护措施是否依然有效,并确认修复不会引入新的风险。随着系统持续变化,网络安全正是以此强化机器人安全。

译文由英文原文机器翻译,术语按行业惯例统一,可在文末展开原文核对。

查看英文原文(28 Zero-Day Findings in Under Two Days. Is Robotics Security Keeping Pace?)

AI Finds 28 Zero-Day Flaws In Physical Robotics Software - IEEE Spectrum

28 Zero-Day Findings in Under Two Days. Is Robotics Security Keeping Pace?

FOR THE TECHNOLOGY INSIDER

FOR THE TECHNOLOGY INSIDER

IEEE Spectrum is the flagship publication of the IEEE — the world’s largest professional organization devoted to engineering and applied sciences. Our articles, videos, and infographics inform our readers about developments in technology, engineering, and science.

Enjoy more free content and benefits by creating an account

Saving articles to read later requires an IEEE Spectrum account

The Institute content is only available for members

Downloading full PDF issues is exclusive for IEEE Members

Downloading this e-book is exclusive for IEEE Members

Access to Spectrum 's Digital Edition is exclusive for IEEE Members

Following topics is a feature exclusive for IEEE Members

Join the world’s largest professional organization devoted to engineering and applied sciences and get access to all of Spectrum’s articles, archives, PDF downloads, and other benefits. Learn more about IEEE →

Join the world’s largest professional organization devoted to engineering and applied sciences and get access to this e-book plus all of IEEE Spectrum’s articles, archives, PDF downloads, and other benefits. Learn more about IEEE →

Access Thousands of Articles — Completely Free

Create an account and get exclusive content and features: Save articles, download collections, and post comments — all free! For full access and benefits, subscribe to Spectrum .

28 Zero-Day Findings in Under Two Days. Is Robotics Security Keeping Pace?

VicOne’s AI-assisted research reveals what faster zero-day discovery means for robot safety

VicOne is a Physical AI cybersecurity company protecting vehicles and robots from cyber threats that can alter behavior and compromise safety.

Humanoid and dog robots watch a massive swirling storm over a modern city walkway

VicOne used AI to find 28 zero-day vulnerabilities in Physical AI tools —these weaknesses can impact robot training, testing, and real-world control.

This article is brought to you by VicOne.

VicOne identified an initial set of 28 zero-day findings in less than two days using AI-assisted research across software and tools supporting the Physical AI ecosystem. These include projects developed or maintained by Google, NVIDIA, and Meta, among others.

Based on initial validation, all 28 findings are classified as zero-day vulnerabilities. The count may change following vendor and vulnerability-authority review, and some findings do not yet have CVE tracking numbers.

Many of these tools help teams train, simulate, test, and operate robots. Weaknesses in them could influence the models teams build, the test results they trust, and the instructions sent to robots.

These findings create two demands for robotics teams: responding as AI helps uncover weaknesses faster, and assessing how each one could affect robot safety and operations.

What AI changes about vulnerability discovery

Vulnerability discovery involves examining code, tracing how systems communicate, and testing whether a suspected weakness can actually be exploited. AI can accelerate parts of this work, helping researchers analyze more software in less time. Attackers could use similar capabilities.

When the weakness is a zero-day — previously unknown and without an available fix — faster discovery can leave teams with little time to respond. While that exposure does not necessarily mean an attack has occurred, teams still need to determine whether they use the affected software and what an attacker could access or change.

Where the findings sit across the robot lifecycle

VicOne’s findings span three stages of the robotics lifecycle: development and training, simulation and validation, and deployment and runtime. Weaknesses at each stage could affect what a robot learns, how it is tested, and how it is controlled.

Bar chart: 50% development, 25% simulation, 25% deployment in robotics workflow.Figure 1. In this research set, 75% of findings involved development, training, simulation, or validation tools, highlighting risks that can arise before a robot is deployed.VicOne

The following scenarios show how a compromised component could influence the robot at each stage.

Development and training: Hidden changes can reach the finished robot
A team downloads a pretrained model to adapt for its robot. Loading a malicious file could give an attacker access to the training system, allowing changes to training data or the model itself. A manipulated model could behave normally during routine testing but respond incorrectly to a particular object, scene, or command after deployment. The risk can enter through a tool the team trusts, before anyone attacks the robot directly.

Simulation and validation: A passing test could give false confidence
Importing a compromised scene or robot model could give an attacker access to the simulation environment. With sufficient control, the attacker could alter simulated sensor readings, physical conditions, or test results. A robot could then appear to pass tests that should have exposed unsafe behavior, leading the team to approve it for real-world use on the basis of misleading evidence.

Deployment and runtime: Operators could lose visibility or control
An attacker could exploit weaknesses to change files that influence motion planning or send malicious map data that crashes the operator’s monitoring display. Compromising a remote-operation computer could also put the instructions sent to the robot at risk. Depending on the system and its safeguards, these weaknesses could alter robot movement or leave operators unable to see what is happening and respond promptly.

These are potential consequences and not demonstrated outcomes for every finding. Actual effects depend on the attacker’s access, the system’s design, and whether other safeguards prevent the changes from reaching physical behavior.

In a separate demonstration, VicOne researchers provide a concrete example of a weakness affecting a protective function. The researchers exploited a vulnerability to prevent a robot’s remote-stop command from working. The case shows why teams need to test whether safeguards remain effective when the software supporting them is compromised.

Video 1. VicOne researchers demonstrate a vulnerability that prevents a robot’s remote-stop command from working.VicOne

A published patch still needs to be proven safe

AI can shorten the search for weaknesses, but a robotics team still needs to determine whether a fix is safe to deploy. Even after a software maintainer releases a patch, teams must identify affected systems, check compatibility, and verify that the update preserves intended behavior.

An earlier successful test remains useful evidence. However, a newly discovered weakness may reveal conditions that test never examined.

Three questions can help teams respond:

Where do we use the affected software? Include training computers, simulation environments, and deployed robots.

What could the weakness change? Assess whether it could affect models, test results, operator visibility, or robot behavior.

How do we verify protection? Confirm that the fix addresses the weakness and that the updated system still operates as intended.

When immediate updating is impractical, teams may need temporary safeguards, such as restricting access to an affected service, while a validated fix is prepared.

From discovery to safer deployment

As details from the 28 findings become available through coordinated disclosure, the next step is for robotics manufacturers is to connect each finding to its potential effect on development, deployment, and robot behavior.

That work also requires understanding how threats could affect robot behavior and how to protect systems already in operation. VicOne Radeis helps teams identify potential zero-day vulnerabilities, assess AI model and system security, and use simulation to evaluate how these risks could affect robot behavior before deployment. VicOne Rthena supports threat detection and responses within approved policies during operation.

Diagram of AI robot security: development, simulation, and operational protection stepsFigure 2. Cybersecurity strengthens robot safety evidence throughout the lifecycle. VicOne helps teams test safety limits against manipulation before deployment and monitor the conditions behind safe behavior in operation.VicOne

Faster vulnerability discovery matters when teams can turn findings into action: identify what needs attention, verify that safeguards still hold, and confirm that a fix does not introduce new risks. That is how cybersecurity strengthens robot safety as systems continue to change.

原文为英文,正文由自动抓取并保留原文。

本文由采集器自动同步

原文发布于 IEEE Spectrum AI(28 Zero-Day Findings in Under Two Days. Is Robotics Security Keeping Pace?),标题、摘要与正文由机器翻译,术语按行业惯例统一,未经人工逐句校订。

阅读原文 ↗