返回首页

对比阅读

对比阅读:Client asks: 'Is this AI?' Only 2% of us admit it 与 客户突然问你方案是不是 AI 写的,目前只有 2% 同行敢说真话

AEN
AI 透明度客户沟通自由职业·

Client asks: 'Is this AI?' Only 2% of us admit it

Last Wednesday at 8 PM, my friend Ailin got blindsided in our group chat

Last Wednesday at 8 PM, my friend Ailin broke down in our group—a client flat-out asked if her proposal was AI-written. She froze for three seconds, then mumbled something about "referencing industry materials." I know that panic. When I landed my first coaching client, they asked me point-blank: "Did you use ChatGPT to draft this?" I instinctively denied it and regretted it the whole night. The truth is, clients fear being hidden from—not the fact that you used AI.

This isn't just Ailin's problem—the whole industry is faking it

a16z (Andreessen Horowitz, the VC fund most aggressive on AI) published a survey last month: they reviewed public materials from 100+ companies on AI usage, and only 2% honestly wrote out "here's what AI does for us." The other 98% either pretended they don't use it, or buried the answer in vague corporate-speak.

But here's the twist: a batch of small teams is doing the opposite. A small NYC design studio I follow now appends a line to every deliverable: "AI-assisted initial draft, 40% human-edited." An SEO micro-team put a full AI policy doc right on their About page. Founder Jamie tweeted: "Clients trust me more now—because I'm not pretending."

You can copy this today

  • Money: $0
  • Time: 30 minutes to write your own "AI usage statement"
  • Technical barrier: zero—if you can type, you can do this
  • First step: open your most recent delivered project and write down which parts you used AI for, which parts you did yourself, and what you changed at the end

My mistake was writing it too "official"—5 bullet points nobody read. Then I copied Jamie's style: three plain-English paragraphs. What I use AI for. What I won't let AI touch. Who's responsible if something breaks.

Advice by stage

Just starting out (no clients yet): Don't sweat this now. But build the muscle anyway—every time AI helps you write something, jot a quick note: "this part was AI-generated, I changed X." Future-you will thank present-you.

Got 1–2 clients: Strongly suggest writing your AI statement this week. Keep it short—three paragraphs is plenty. My template: "I use AI for first drafts and research, but every deliverable gets human review; client data never gets fed into public AI tools; tell me in advance if you'd rather I don't use AI." Send it over—usually they think you're more serious, not less.

Scaling up (team forming): This stops being a personal statement and becomes a "team AI policy." You need rules: which roles can use AI, what client data can't be uploaded, who's on the hook when AI hallucinates. Google "AI policy template for small agency," grab an online doc, tweak it, ship it.

One last thing: not everyone needs this. If your clients never ask, drafting a formal statement just looks performative. But the moment one client asks you "is this AI?"—that night, you'll think of Ailin. Writing it early beats scrambling after the question. Way more comfortable.

BZH
AI 透明度客户沟通自由职业·

客户突然问你方案是不是 AI 写的,目前只有 2% 同行敢说真话

上周三晚上八点,阿琳被客户当场面问了

上周三晚上八点,我朋友阿琳在群里崩溃——客户直接问她方案是不是 AI 写的,她愣了三秒,编了句"借鉴了行业资料"。我懂那种慌。我自己刚接第一个私教客户的时候,也被问过"你是不是用 ChatGPT 写文案",我下意识否认了,回家后悔了一晚上。其实客户更怕的是你藏着掖着,不是你用了 AI。

这事不是阿琳一个人的问题,是整个行业都在装

a16z(就是那个硅谷最敢投 AI 的基金 Andreessen Horowitz)上个月发了个调研:他们看了 100 多家公司在公开材料里披露的 AI 使用情况,结果只有 2% 老老实实写出来"我们用 AI 干了啥"。剩下 98% 要么假装没用,要么写得云里雾里。

但有意思的是,已经有一批小团队反着来。我关注的一个纽约小设计工作室,每份交付稿最后都加一句"AI 辅助生成初稿,人工调整 40%";做 SEO 的一个小团队直接在自己官网介绍页放了一份 AI 政策文档。创始人 Jamie 在推特上说:"客户反而觉得我靠谱,因为我没装。"

你今天就可以抄这个做法

  • 钱:0 元
  • 时间:30 分钟写一份你自己的"AI 使用声明"
  • 技术门槛:不需要懂任何代码,会打字就行
  • 第一步:打开你最近交付的一个项目,对着它写下"哪些环节我用了 AI、哪些环节我自己干的、最终我又改了什么"

我之前犯过的错是写得太"正规",列了 5 条原则客户根本不看。后来我就学 Jamie——大白话写三段话:我用 AI 干啥、我不让 AI 干啥、出了问题我负责。

分人群建议

如果你刚起步(还没客户):这事现在不用急。但你可以现在就养成习惯——每次用 AI 帮你写东西,顺手备注一句"这段 AI 生成的,我改了哪里"。未来你会感谢自己。

如果你有 1-2 个客户:强烈建议本周就写一份你的 AI 声明,不要长,三段话够了。我自己用的模板是"我用 AI 写初稿/做调研,但所有交付内容我都人工过一遍;客户数据不喂给公网 AI 工具;如果你不要我用 AI,提前说"。发给客户看,通常他会觉得你认真。

如果你在扩规模(团队开始有人):这事就变成了"团队 AI 政策",不是个人声明了。你需要规定:哪些岗位可以用 AI、哪些客户数据不能上传、出现 AI 幻觉谁负责。建议直接搜一份"AI 政策模板"(搜 "AI policy template for small agency"),抄一份在线文档改改就用。

最后说一句:这做法不是所有人都需要。如果你客户从来没问过,你硬要写个声明反而显得做作。但只要有一个客户问过你"是不是 AI",那天晚上你就会想起阿琳——早点写,比被问到再写,舒服得多。