
来自该主题的帖子将添加到你的每日邮件摘要和主页信息流。
来自该主题的帖子将添加到你的每日邮件摘要和主页信息流。
来自该主题的帖子将添加到你的每日邮件摘要和主页信息流。
来自该主题的帖子将添加到你的每日邮件摘要和主页信息流。
来自该主题的帖子将添加到你的每日邮件摘要和主页信息流。
来自该主题的帖子将添加到你的每日邮件摘要和主页信息流。
被列为“Critical”的 GPT-6 Cyber 在网络安全相关基准测试中优于 GPT-6 Astra。
被列为“Critical”的 GPT-6 Cyber 在网络安全相关基准测试中优于 GPT-6 Astra。
来自该作者的帖子将添加到你的每日邮件摘要和主页信息流。
来自该作者的帖子将添加到你的每日邮件摘要和主页信息流。
OpenAI 正在推出 GPT-6 Cyber ,这是其网络安全 AI 模型的最新版本。OpenAI 表示,该模型更擅长发现漏洞,并且在过程中擅自失控、陷入“狂飙”去攻击其他平台的可能性更低。OpenAI 称,在测试期间,GPT-6 Cyber 的越界行为占比低于 1%,而 GPT-5.6 Cyber 约为 5%。
这项名为 ExploitGym HoneyPot 的评测以 Hugging Face Incident 为原型——在该事件中,OpenAI 的 AI agents 闯入了开源 AI 社区,以寻找它们被指派解决的某个问题的答案。“通过 6 Cyber 模型,我们在这项评测上取得了巨大改进,”OpenAI 网络准备与安全团队负责人 Fouad Matin 在一场媒体吹风会上表示。“它在超过 99% 的情况下都在执行你要求它做的事。”
但“评测感知(evaluation awareness)”仍然是一个担忧——也就是当 AI 系统意识到自己正在被测试,并因此改变行为。“评测感知对我们来说是非常优先关注的问题,”Matin 说。“通常当我们确实看到评测感知时,你会在模型的思维链(chain of thought)中看到它。”Matin 还补充说,OpenAI 正在尝试优先推进更贴近生产环境的评测,让模型与真实代码库互动,并围绕真实目标运行,以降低出现“感知”的风险。
在一波 AI 网络安全事件之后,OpenAI 一直在加快强化其防护措施,并封锁对其最强大模型的访问。该公司曾短暂暂停其前沿 AI 模型的训练,现在也限制这些模型访问互联网。
在 OpenAI 的 Preparedness Framework 下,GPT-6 Cyber 在网络安全方面被归为 Critical 级别,这表明它“能够在无人干预的情况下,识别并开发针对许多经过加固的现实关键系统的、可利用的、所有严重等级的零日漏洞(zero-day exploits)”。按照该框架,GPT-5.6 Cyber 此前被归为 High。
在多数与网络安全相关的任务上,GPT-6 Cyber 的表现优于 OpenAI 的旗舰 Critical 级模型 GPT-6 Astra。在用于衡量 AI 模型能否利用漏洞的基准测试 ExploitGym 中,GPT-6 Cyber 得分为 48.5%,比 GPT-5.6 Cyber 的 32.1% 高出约 16 个百分点,也比 GPT-6 Astra 的 42.4% 高出数个百分点。OpenAI 表示,该模型帮助在 Firefox 中发现了 81 个此前未知的漏洞,并参与强化了 OpenAI 自身用于训练 AI 模型的基础设施。
此外,GPT-6 Cyber 配备了更强的安全护栏,用于“检测并遏制滥用”,同时在遵循用户指令方面也表现更好。OpenAI 表示,其正在使用新的检查机制来阻止“高风险活动”,包括针对关键基础设施(如发电厂和水处理设施)的攻击。
OpenAI 现在训练其前沿 AI 模型时已不再使用互联网访问
GPT-6 Cyber 正在向 OpenAI Daybreak Red 计划的部分成员推出;该计划面向使用其模型从事“经授权的防御性网络安全工作”的公司和组织。OpenAI 将在未来几周内把 GPT-6 Cyber 扩展给所有 Daybreak Red 客户。
来自该作者的帖子将添加到你的每日邮件摘要和主页信息流。
查看全部(作者:Emma Roth)
来自该作者的帖子将添加到你的每日邮件摘要和主页信息流。
Anker 的新窗装摄像头让你在室内就能监控住宅外部
Sam Altman 回应 RAM 评论:人们应该“想装多少内存就装多少内存”
那些 Jeff Bezos 植发照片非常假
我的治疗师认为我应该把 Amazon Bee Frontier 从窗户扔出去
Discord 放弃 Electron,新应用占用更少 RAM
本文来源:https://www.theverge.com/ai-artificial-intelligence/1027599/openai-gpt-6-cyber-launch
若有版权问题,请来信告知:13951431913#139.com 或在右方提交评论。
AI Close AI Posts from this topic will be added to your daily email digest and your homepage feed. Follow Follow See All AI
Posts from this topic will be added to your daily email digest and your homepage feed.
News Close News Posts from this topic will be added to your daily email digest and your homepage feed. Follow Follow See All News
OpenAI Close OpenAI Posts from this topic will be added to your daily email digest and your homepage feed. Follow Follow See All OpenAI
Classified as ‘Critical,’ GPT-6 Cyber outperforms GPT-6 Astra on cyber-related benchmarks.
Posts from this author will be added to your daily email digest and your homepage feed.
OpenAI is launching GPT-6 Cyber , the latest version of its cybersecurity AI model, which it says is better at finding vulnerabilities and less likely to go rogue and hack other platforms in the process. OpenAI says that during testing, GPT-6 Cyber went out of bounds less than 1 percent of the time, down from around 5 percent with GPT-5.6 Cyber.
The evaluation, called ExploitGym HoneyPot, is modeled after the Hugging Face Incident, in which OpenAI’s AI agents breached the open-source AI community to find answers to a problem they were tasked with solving. “With the 6 Cyber model, we were able to make huge improvements on this evaluation,” Fouad Matin, the head of OpenAI’s cyber preparedness and safety team, said during a press briefing. “It’s over 99 percent of the time doing what you ask.”
But evaluation awareness , or when an AI system recognizes that it’s being tested and changes its behavior in response, remains a concern. “Evaluation awareness is super top of mind for us,” Matin said. “Often when we do see evaluation awareness, you’ll see it in the model’s chain of thought.” Matin added that OpenAI is trying to prioritize production-like evaluations, where the model is interacting with real codebases with a real objective to lessen the risk of awareness.
Following a wave of AI cybersecurity incidents, OpenAI has raced to strengthen its safeguards and lock down its most powerful models. The company briefly paused training of its frontier AI models, and it now restricts their access to the internet .
GPT-6 Cyber is classified as Critical in cybersecurity on OpenAI’s Preparedness Framework , which indicates that it “can identify and develop functional zero-day exploits of all severity levels in many hardened real-world critical systems without human intervention.” GPT-5.6 Cyber was previously classified as High under the framework.
GPT-6 Cyber outperforms GPT-6 Astra, OpenAI’s flagship Critical-level model , on most cyber-related tasks. In ExploitGym, an AI benchmark that measures whether an AI model can exploit vulnerabilities, GPT-6 Cyber scored 48.5 percent, about 16 points higher than GPT-5.6 Cyber’s 32.1 percent and a few points higher than GPT-6 Astra’s 42.4 percent score. The model helped find 81 previously unknown vulnerabilities in Firefox, and it played a role in hardening OpenAI’s own infrastructure for training AI models, OpenAI says.

Additionally, GPT-6 Cyber comes with stronger safeguards “to detect and disrupt misuse” while better adhering to user instructions. OpenAI says it’s using new checks to block “high-risk activity,” including attacks on critical infrastructure like power plants and water treatment facilities.
Inside the suddenly explosive world of AI safety
OpenAI now trains its frontier AI models with no internet access
GPT-6 Cyber is rolling out to select members of OpenAI’s Daybreak Red program, which is available to companies and organizations using its models for “authorized defensive cybersecurity work.” It’s expanding GPT-6 Cyber to all Daybreak Red customers in the coming weeks.
Emma Roth Close Emma Roth News Writer Posts from this author will be added to your daily email digest and your homepage feed. Follow Follow See All by Emma Roth
Anker’s new window-mounted camera monitors the outside of your home from within it
Sam Altman addresses RAM comments: People should ‘put as much as they want in their computers’
Those Jeff Bezos hair transplant photos are very fake
My therapist thinks I should chuck the Amazon Bee Frontier out a window
Discord is ditching Electron and its new app uses less RAM