当初创公司请求免费安全服务时

3作者: hdue3 个月前原帖
几周前,我探索了[已编辑],这是一个获得YC支持的AI后端平台。像许多安全研究人员一样,我倾向于测试新工具,以观察它们如何处理常见的攻击向量。 ## 漏洞 *授权缺陷*: [已编辑] 限制免费用户只能使用3个项目,更多则需付费。但他们的API并没有强制执行这一限制。任何人都可以绕过前端,直接调用API。 这个经典的缺陷意味着免费用户可以生成无限内容,付费用户的价值下降,商业模式也随之崩溃。 *用户体验问题*:该平台的导航混乱,设计不一致,层次结构差,工作流程笨拙,入门指导不清。当产品体验如此粗糙时,安全缺陷只是忽视的另一个标志。 ## 响应 我在他们的社区频道询问了他们的披露流程。创始人回复道: “嗨 [姓名],我刚看到你在公共频道的消息。目前我们不在招聘,但有人在帮助改善平台,这对未来我们招聘时是个很好的测试。如果你想贡献,随时可以向我们报告漏洞或安全问题。如果是安全相关的,最好通过私信而不是公共频道。” 翻译:*请为我们免费做安全工作。也许有一天我们会雇用你。* ## 我为何不披露 我没有透露细节,因为: - 没有漏洞奖励或认可系统 - 安全研究被框架为“免费测试” - 对未来考虑的模糊承诺,而不是当前的补偿 - 没有披露政策或时间表 - 整体缺乏专业性 发现并负责任地报告漏洞需要技能。期望研究人员免费为之工作,尤其是来自一家获得资金支持的初创公司,这是不可接受的。 ## 更广泛的问题 这反映了一个更大的初创公司问题:希望获得社区的帮助却不愿支付。公司在筹集数百万资金的同时,常常要求无偿的质量保证、安全审计、漏洞报告和用户体验反馈。 ## 优秀公司的做法 优秀的公司通常具备: - 明确的披露政策和定义的时间表 - 漏洞奖励计划(即使是小型的也表示尊重) - 与研究人员的专业沟通 - 对负责任披露的公开认可 这并不需要太多。即使是10美元的礼品卡和一句感谢也很重要。 ## 当前状态 一个月后,漏洞仍未修复,用户体验依然粗糙。 对于用户来说,这意味着不准确的使用跟踪、破裂的经济模型、可能更深层次的问题以及持续的挫败感。对于公司而言,这揭示了一种文化,安全、用户体验和尊重都是事后考虑的事情。 ## 对创始人的教训 *安全基础*: - 在服务器端强制执行所有限制。绝不要信任前端。 - 发布简单的披露政策。 - 尊重研究人员,我们是在试图帮助。 *文化基础*: - 不要要求免费劳动。 - 将反馈视为有价值的,而不是免费的质量保证。 - 记住,第一印象是持久的。 安全社区希望提供帮助,但不希望以贬低专业知识为代价。 构建安全的产品。创造直观的体验。尊重那些帮助你改进的人。安全债务迅速累积,但用户体验债务则更快地扼杀用户采用。 --- 你是否也有类似的经历,AI初创公司期望免费提供安全工作?你是如何处理那些忽视安全的公司的?
查看原文
A few weeks ago, I explored [redacted], a YC-backed AI backend platform. Like many security researchers, I tend to poke at new tools to see how they handle common attack vectors.<p>It didn’t take long to find issues, both in security and user experience.<p>## The Vulnerabilities<p>*Authorization Flaw*: [redacted] limits free users to 3 items, with a paywall for more. But their API doesn’t enforce this. Anyone can bypass the frontend and call the API directly.<p>This classic flaw means free users can generate unlimited content, paid tiers lose value, and the business model collapses.<p>*UX Problems*: The platform also has confusing navigation, inconsistent design, poor hierarchy, clunky workflows, and unclear onboarding. When the product experience feels this raw, security flaws are just another sign of neglect.<p>## The Response<p>I asked in their community channel about their disclosure process. The founder replied:<p>“hi [name], i just saw your message on the general channel. right now, we are not hiring, but people are helping improving the platform and this is a good test for the future, when we will hire people. if you want to contribute, feel free to report bugs or security issues to us. if security related, it&#x27;s best on private dms rather than on general channel”<p>Translation: <i>Please do free security work for us. Maybe we’ll hire you someday.</i><p>## Why I Didn’t Disclose<p>I withheld details because: - No bug bounty or acknowledgment system - Security research framed as &quot;free testing&quot; - Vague promise of future consideration, not present compensation - No disclosure policy or timeline - Overall lack of professionalism<p>Finding and responsibly reporting vulnerabilities takes skill. Expecting researchers to do it for free, especially from a funded startup, is unacceptable.<p>## The Broader Problem<p>This reflects a larger startup issue: wanting community help without paying for it. Companies routinely ask for unpaid QA, security audits, bug reports, and UX feedback while raising millions.<p>## What Good Companies Do<p>The best companies have: - Clear disclosure policies with defined timelines - Bug bounty programs (even small ones show respect) - Professional communication with researchers - Public acknowledgment for responsible disclosure<p>It doesn’t take much. Even a $10 gift card and a thank-you matter.<p>## Current Status<p>A month later, the vulnerability is still unfixed, and UX remains rough.<p>For users, this means inaccurate usage tracking, broken economics, possible deeper issues, and ongoing frustration. For the company, it reveals a culture where security, UX, and respect are afterthoughts.<p>## Lessons for Founders<p>*Security basics*: - Enforce all limits server-side. Never trust the frontend. - Publish a simple disclosure policy. - Respect researchers, we’re trying to help.<p>*Cultural basics*: - Don’t ask for free labor. - Treat feedback as valuable, not free QA. - Remember that first impressions last.<p>The security community wants to help, but not at the cost of undervaluing expertise.<p>Build secure products. Create intuitive experiences. Respect those who help you improve. Security debt compounds quickly, but UX debt kills adoption even faster.<p>---<p>Have you had similar experiences with AI startups expecting free security work? How do you handle companies that dismiss security?