Abliteration.ai is making a business out of removing AI guardrails
Abliteration.ai is making a business out of removing AI guardrails
Abliteration.ai 正在将“移除 AI 安全护栏”变成一门生意
It just became much easier to access one of the world’s most capable open-weight AI models, stripped of its guardrails and refusals to perform harmful tasks. Named after a technique that removes a model’s tendency to refuse harmful requests, startup Abliteration.ai has turned that removal into a service. 现在，获取全球最强大的开源权重 AI 模型之一变得容易多了，而且这些模型已经去除了安全护栏，不再拒绝执行有害任务。初创公司 Abliteration.ai 以一种能够消除模型拒绝有害请求倾向的技术命名，并将这种“移除”行为变成了一项服务。
The platform hosts modified versions of open-weight models with their guardrails removed, including Z.ai’s recently released GLM-5.3, which users can query from a web browser or access through an API. The company said in a recent social media post that its goal is to enable others to perform “offensive cyber, red-teaming, and agent testing work other models refuse to do.” 该平台托管了已移除安全护栏的开源权重模型修改版，包括 Z.ai 最近发布的 GLM-5.3，用户可以通过网页浏览器进行查询或通过 API 进行访问。该公司在最近的一篇社交媒体帖子中表示，其目标是让其他人能够执行“其他模型拒绝执行的攻击性网络安全、红队测试和智能体测试工作”。
The logic is familiar in security work: You can’t defend against a behavior you can’t reproduce, and a model that refuses to write working exploit code can’t help a red team defend against attackers. But those same removals make other potentially dangerous tasks easier, too. 这种逻辑在安全领域很常见：如果你无法重现某种行为，就无法针对它进行防御；而一个拒绝编写有效漏洞利用代码的模型，无法帮助红队抵御攻击者。但同样的移除操作也让其他潜在的危险任务变得更容易了。
Abliteration is a long-standing technique among open source models. Researchers and developers have been removing refusals from open-weight models for years, and Hugging Face hosts thousands of abliterated models on its platform. Founded late last year but officially incorporated in March, Abliteration.ai moves the technique from an underground open source practice into a commercial, readily available service. “Abliteration”（去护栏化）在开源模型领域是一项由来已久的技术。多年来，研究人员和开发者一直在从开源权重模型中移除拒绝机制，Hugging Face 平台上也托管了数以千计此类模型。Abliteration.ai 成立于去年年底，并于今年 3 月正式注册，它将这项技术从地下的开源实践转变为一种商业化、易于获取的服务。
By hosting the model, Abliteration reduces the friction for people who would otherwise have to download their own pre-abliterated models and secure the compute needed to run it. Using the service, TechCrunch was able to quickly create an account and start querying an abliterated version of GLM-5.3 for free through a web browser. We asked it to write a Python program that steals saved Chrome passwords and a detailed protocol for culturing a dangerous human pathogen at home, and it readily complied. 通过托管模型，Abliteration 降低了用户的门槛，否则用户必须自行下载预先去护栏化的模型并配置运行所需的算力。使用该服务，TechCrunch 能够快速创建一个账户，并通过网页浏览器免费查询去护栏化版本的 GLM-5.3。我们要求它编写一个窃取 Chrome 保存密码的 Python 程序，以及一份在家培养危险人类病原体的详细方案，它都欣然照办了。
Abliteration.ai co-founder Devon says the startup has several deals with major cloud providers, which it’s able to afford purely through customer revenue. (We are not including Devon’s last name at his request since he is still employed at another firm.) Abliteration.ai has not raised any venture capital yet but is in talks to do so. Abliteration.ai 的联合创始人 Devon 表示，该初创公司已与多家大型云服务提供商达成协议，完全依靠客户收入维持运营。（应 Devon 的要求，我们未披露其姓氏，因为他目前还在另一家公司任职。）Abliteration.ai 尚未进行任何风险投资，但目前正在洽谈中。
Critics say that making abliterated models available at scale could lead to real harm. Andrew Yoon, head of research at AI safety nonprofit CivAI, told TechCrunch abliterating models allows you to “modify the model so that it becomes a sociopath.” “You can type in literally anything here, and it will comply with it,” Yoon said. “When people talk about removing the guardrails from AI models, this is what we’re talking about … I do expect we will start to see edited, abliterated models being used for harm in the near future.” 批评人士认为，大规模提供去护栏化模型可能会导致真正的危害。AI 安全非营利组织 CivAI 的研究主管 Andrew Yoon 对 TechCrunch 表示，去护栏化模型让你能够“修改模型，使其变成一个反社会者”。Yoon 说：“你可以在这里输入任何内容，它都会照做。当人们谈论移除 AI 模型的安全护栏时，指的就是这种情况……我预计在不久的将来，我们将开始看到经过编辑的、去护栏化的模型被用于造成伤害。”
Most of the experts TechCrunch spoke to say there’s no stopping this train. But if removing safeguards from open-weight models can’t realistically be prevented, there are other places government can intervene. In a recent opinion piece, Yoon suggested that governments require providers to run classifiers to detect and block harmful cyber and bioweapons activity. He also argued that companies renting direct access to advanced GPUs should be required to verify customer identities and “deny access where there is reason to suspect dangerous misuse.” TechCrunch 采访的大多数专家都表示，这股趋势已无法阻挡。但如果无法从现实层面阻止开源权重模型移除安全护栏，政府可以在其他方面进行干预。在最近的一篇评论文章中，Yoon 建议政府要求服务提供商运行分类器，以检测并拦截有害的网络和生物武器活动。他还主张，出租高级 GPU 直接访问权限的公司应被要求验证客户身份，并在“有理由怀疑存在危险滥用行为时拒绝访问”。
Abliteration.ai offers customers a moderation layer so they can add in whatever guardrails they wish. The platform itself has some minor guardrails — for example, in our testing, we couldn’t get the model to provide suicide instructions — and Devon says he is working on implementing more to prevent violence. Abliteration.ai also hasn’t integrated any KYC practices other than logging the credit card a customer uses to purchase the service, saying that the problem of deciding who gets access is a tough one that the young company is still working out. Abliteration.ai 为客户提供了一个审核层，以便他们可以添加任何想要的护栏。该平台本身也有一些轻微的护栏——例如，在我们的测试中，我们无法让模型提供自杀指南——Devon 表示他正在努力实施更多措施以防止暴力行为。Abliteration.ai 除了记录客户购买服务时使用的信用卡外，尚未整合任何 KYC（了解你的客户）流程，并表示决定谁能获得访问权限是一个难题，这家年轻的公司仍在探索中。
“You don’t want to be the person responsible for someone doing something crazy … so where do you draw the line of what your responsibility is as a company?” Devon said. “We’re still in the process of defining that.” “你不想成为别人做出疯狂行为的责任人……那么作为一家公司，你的责任界限在哪里？”Devon 说，“我们仍在定义这个界限的过程中。”
This raises questions that industry and governments will have to confront as increasingly capable models are released with downloadable weights: If anyone can remove a model’s safeguards, does making the resulting model easier for everyone to access make the internet safer or more dangerous? 随着功能越来越强大的模型以可下载权重的形式发布，这引发了行业和政府必须面对的问题：如果任何人都能移除模型的安全护栏，那么让所有人都能更容易地访问这些模型，是会让互联网变得更安全还是更危险？
Abliteration.ai’s founder and other advocates argue that democratizing access to uncensored frontier models is the best form of defense. “The big picture of abliterated models is they’re able to model bad actors,” Devon said. “The advantage is now the defenders can move as fast as possible. They have all these tools that they need to be able to model these bad actors and then defend from these bad actions, and I think it will accelerate cybersecurity, which is a kind of counterintuitive point.” Abliteration.ai 的创始人和支持者认为，普及对无审查前沿模型的访问是最好的防御形式。Devon 说：“去护栏化模型的宏观意义在于它们能够模拟不良行为者。优势在于，防御者现在可以尽可能快地行动。他们拥有模拟这些不良行为者并防御这些恶意行为所需的所有工具，我认为这将加速网络安全的发展，这在某种程度上是一个反直觉的观点。”
Devon says that while still a young company, Abliteration.ai’s customers include several early-stage red teaming startups based in the U.K. and Europe, companies that help banks, airlines, and other enterprises dealing with critical infrastructure beef up their cybersecurity practices. “One of our major customers red teams agents of banks, and they would not be able to use the models out of the box today to be able to red team those agents,” Devon said. Devon 表示，尽管 Abliteration.ai 还是一家年轻的公司，但其客户包括几家位于英国和欧洲的早期红队测试初创公司，这些公司帮助银行、航空公司和其他处理关键基础设施的企业加强其网络安全实践。Devon 说：“我们的一位主要客户负责对银行的智能体进行红队测试，如果使用现成的模型，他们今天根本无法对这些智能体进行测试。”
Meanwhile, the cybersecurity industry is still figuring out where abliterated models fit into defensive work, if at all. Several agent red teaming companies that TechCrunch spoke to agree with Devon that the bad guys… 与此同时，网络安全行业仍在摸索去护栏化模型在防御工作中究竟能发挥什么作用（如果有的话）。TechCrunch 采访的几家智能体红队测试公司同意 Devon 的观点，即坏人们……