# Firm AI Objection Answer

> Prepares an honest answer to "why can't AI do this?", with the client's words and the question behind them, the engagement split into tasks AI can do well, tasks AI helps a person do and tasks that need a person and why, what you say in order, and an option where the client does part with their own AI. Use for "run firm-ai-objection-answer", "the client says AI could do this", "why can't AI just do it", "what am I paying you for when I have Claude", "answer the AI objection", "justify my fee against AI", "prepare me for the AI question", part of the Claude for Consultants · Team in Claude Pack by 90n.

- Skill: `polar-bear-org/firm-ai-objection-answer` (Agent Skill)
- Install (CLI): `npx get-skill add polar-bear-org/firm-ai-objection-answer`
- Raw SKILL.md: https://api.skillmd.com/api/skills/polar-bear-org/firm-ai-objection-answer/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: polar-bear-org (https://skillmd.com/u/polar-bear-org)
- Updated: 2026-11-14
- Page: https://skillmd.com/skills/polar-bear-org/firm-ai-objection-answer

---


# AI Objection Answer

## When To Use
It usually arrives lightly, almost as a joke: "couldn't we just ask Claude?" Your pulse goes up and you start listing things AI gets wrong. Do not. Run this before the next conversation, to work out which parts of this engagement AI really can do, which parts need you, and what that should mean for scope and price.

## When Not To Use
If the AI remark is cover for price, timing or "we will handle it internally", the Objection Handling Map fits better. If the client asks in writing how you use AI yourself, send your AI Use Statement.

## Inputs
- What the client said, as near to word for word as your notes allow, and at what point
- Your scope or proposal, split into the tasks you would really carry out
- Your AI use statement, or a few lines on how you use AI on work like this
If you have none of this, I start from the client's sentence and one line of scope, and mark the answer as a first draft.

## Approach
Two methods are combined. The task-level view of the Anthropic Economic Index (anthropic.com/economic-index) looks at work one task at a time and separates AI doing a task from AI helping a person do it; it is used as a way of looking, with no figures. LAER, from Carew International, gives the conversation its order: listen, acknowledge, explore, respond. The failure it prevents is the defensive answer that claims AI cannot do something the client watched it do last night. You lose the argument and their confidence together.

## Workflow
1. Ask, in one go: what exactly did the client say, and who said it? What have they already tried with AI? What does this piece of work cost them as quoted?
2. Listen. Write the objection in their words, then the question underneath: "are you worth this fee", "will it really take that long", or "could we do it ourselves"? You know the room; you pick.
3. Acknowledge. One sentence agreeing with what is true. If AI can produce the first analysis or the first draft, say so without a "but".
4. Split the engagement into its real tasks, in three columns: AI can do it well; AI helps a person do it; needs a person, with the reason (judgment, context only the client's people hold, accountability, the risk if it goes wrong). Fill the first column first, and never shrink it to protect the fee.
5. Explore. Draft one or two questions that reveal what they expect: what did the AI output give you, what was missing, who would check it before it reaches your board?
6. Respond, in this order: what AI does well here, what stays with you, how you already use AI to their benefit, and a real option in which they run the first column with their own tools while you do the rest, at a price you set. Then cut every line that sounds like a rebuttal.

## Output Format
```markdown
# Answer to the AI Question
**What the client said:** [their words; quotation marks only if from your notes]
**The question behind it:** [fee / time / do it ourselves / other]
## Task split
| Task in this engagement | AI can do it well | AI helps a person do it | Needs a person, and why |
|---|---|---|---|
| [task] | [yes / no] | [yes / no] | [judgment / context / accountability / risk] |
## What I say, in order
1. Acknowledge: [the true part, in one sentence]
2. Explore: [question] / [question]
3. What AI handles well in this work: [plain list]
4. What stays with me: [plain list]
5. Where AI already works for you in my practice: [from your AI use statement]
6. The option with your own AI: [what you do, what I do; price set by me]
## Decision
[You] decide by [date] whether to offer the shared option, and at what price, before [next meeting with client role].
```

## Done When
- The "AI can do it well" column was filled first, and no task was moved out of it to protect the fee
- Every "needs a person" line gives its reason
- At least one explore question is in the answer
- The shared option is real, scoped and priced by you, or you have chosen not to offer it

## Quality Bar
- Agree with what is true before anything else
- No claim that AI cannot do a task unless you have checked
- No running down of AI, of the client's tools or of how they use them
- Your own use of AI is said aloud
- Say what AI does well first; never sell against AI and never hide your own use. I invent no claim about what AI can or cannot do, and you give the answer in person.

## Next
Run firm-quote-comparison (Like-for-Like Quote Comparison) to prepare for the same objection arriving as a cheaper quote.

## About the makers

This pack is made by 90n ([90n.ai](https://90n.ai)): AI-native ways of working and tools for Product, Marketing and HR teams, built by ex-McKinsey founders. We map where time and money leak, redesign the work around AI and stay until your people run it themselves, and we run our own company on Claude. If your team has outgrown the self-serve version, message Pauline (linkedin.com/in/paulinebertry).

