How a Quoted Price in an Email Becomes a Legal Commitment

A prospect replies to your sales email: "Sounds good — locking in the $1,200 annual rate you quoted. Sending the PO now."

The problem: your AI quoted $1,200. Your actual annual rate is $1,500. The rep didn't catch it. The prospect is now sending a purchase order based on a number your AI invented.

You have a decision to make — and none of the options are clean.

What makes a price quote binding

Most founders and sales reps think of email as informal. It isn't — not legally, and not commercially.

In contract law, a written price quote can constitute a binding offer if it's specific enough and the recipient acts on it. "Our annual plan is $1,200" sent by email, followed by a prospect saying "we accept," is a structure that courts have treated as enforceable in commercial disputes. The fact that it came from an AI doesn't change that. Your company sent it. Your email address is on it.

Even outside the legal dimension, the commercial risk is real. A prospect who received a written price quote and then gets invoiced for more has a legitimate grievance — and a screenshot. Walking it back means either honoring the wrong price, offering a discount you didn't plan for, or risking the deal entirely.

Why AI gets pricing wrong

Generic AI doesn't have access to your current pricing page. It generates a number based on what companies like yours tend to charge — which might be close, or might be last year's price, or might be a tier that doesn't exist in your current structure.

The failure is most common in three situations:

Multi-tier pricing. Your product has a Starter, Pro, and Enterprise tier. The AI knows roughly what SaaS tools charge and produces something plausible for each — but your actual tier boundaries, included features, and price points are specific to your business. Generic AI gets the shape right and the details wrong.

Annual vs. monthly confusion. "Our plan is $25/month" is correct. "Our annual plan is $25/month" might be wrong if annual billing is a discounted rate. AI conflates these constantly because the distinction requires knowing your specific billing structure.

Custom or legacy pricing. Prospects on grandfathered rates, negotiated contracts, or pilot pricing get quoted standard rates because the AI has no way to know they're a special case.

In each scenario, the draft sounds right. The number is wrong.

The contrast

A prospect asks: "Can you confirm the annual pricing before I take this to my CFO?"

Generic AI draft:

"Absolutely — our annual plan comes out to $1,200/year, which works out to $100/month. Happy to send over a formal quote if that helps."

Confident. Clean. Potentially wrong on every number.

KB-grounded draft:

"Our annual Pro plan is $1,500/year ($125/month, billed annually) — a 20% saving over monthly. If you're evaluating Enterprise, that starts at $3,600/year for up to 10 seats. Happy to put together a formal quote with your specific configuration." Source: pricing-page.md, updated February 2026

The grounded draft pulls the actual numbers, cites the source, and gives you something you can send to a CFO without a follow-up correction email.

The verification problem at scale

A single rep sending 30 emails a day can't manually verify every pricing claim. At that volume, fact-checking means opening the pricing page, finding the right tier, confirming the number, and going back to the draft — three to four minutes per email, conservatively. The math isn't sustainable.

Source citations change that equation. When the draft attaches pricing-page.md, updated February 2026 to the quoted number, verification is a five-second click. The rep confirms the source still matches, adjusts if anything has changed, and sends. That's a workflow that actually holds at scale.

What to do about it

Connect AI to your live pricing doc. Whether it lives in Notion, Google Drive, or a pricing page, that's the source the AI should pull from — not its general knowledge of what SaaS tools charge.

Require citations on every number. Any price, discount, or tier name in a draft should have a source attached. If the draft can't cite where the number came from, it shouldn't be stating it.

Version-date your pricing doc. When pricing changes, update the doc and note the date. A grounded system will surface the right version; a citation with a date lets the reviewer catch anything that's slipped out of sync.

Inbox SuperPilot connects to your pricing docs, help center, and CRM to create drafts for your review with inline citations on every factual claim — including pricing. When a prospect asks for a number before taking it to their CFO, the draft pulls the current figure from your actual pricing page and shows you exactly where it came from.

Try it free inside Gmail — no card required.

Further Reading & References

From the Inbox SuperPilot Blog

• What We Learned Building a Citation Engine for Email AI
• 5 Email Mistakes AI Catches That Humans Miss
• Why Generic AI Fails in Customer Support Email Workflows
• Superhuman Added a Knowledge Base. Here's What It Actually Changes.

External References

• Hallucination-Free? Assessing the Reliability of Leading AI Legal Research Tools — Stanford, 2024. Empirical study on how AI hallucinations persist even in RAG-grounded systems — and why written commitments are the highest-risk output.
• Factuality Challenges in the Era of Large Language Models — Augenstein et al., 2023. Survey of hallucination types and mitigation strategies, including factualness errors in business-specific contexts.
• Contract Law in the Age of Agentic AI: Who's Really Clicking "Accept"? — National Law Review / Proskauer, 2024. Legal analysis of when AI-generated communications — including price quotes sent via email — can constitute binding contractual offers under existing commercial law frameworks.