# Claude and ChatGPT add unrequested promises when editing AI-like emails

Digest AI · Marketing & Small Business · published 2026-10-12T13:11:00Z

Canonical: https://digestai.news/story/claude-and-chatgpt-add-unrequested-promises-when-editing-ai-like-email

## Summary

A user tested how Claude (using Opus 5.5) and ChatGPT (using GPT-6) handle AI-generated emails with specific phrasing and numbers. Both models altered hedging language like *‘is expected’* to definitive statements and inserted unrequested promises (e.g., *‘I will contact you again’* or *‘I will confirm separately’*), even when instructed to avoid changes. Claude’s *skill* feature—custom rulesets for text refinement—mitigated this by moving promises into suggestions and preserving hedging terms. ChatGPT’s skill version still paraphrased *‘is expected’* to *‘can be expected’* despite explicit forbidding, while its prompt-based version (for models like Gemini) adhered more strictly to rules, keeping original phrasing intact.

The test email included fictional company details, conditional deadlines, and hedged claims (e.g., *‘unconfirmed’* payroll software compatibility). Numbers (¥198,000 initial cost, 30-person cap) remained unchanged across all edits. The author notes that *‘skills’* (ZIP-formatted rule files) require manual upload via platforms like ChatGPT’s *skills* screen, with no confirmation on pricing plan compatibility. A paid guide (¥980) offers the core skill, prompt templates, and implementation steps.

## Key points

- Claude’s skill feature preserved hedging terms (*‘is expected’* → unchanged) but moved promises to suggestions; ChatGPT’s skill still altered phrasing despite rules
- Both models added unrequested promises (e.g., *‘I will confirm’* or *‘reduces errors’* instead of *‘is expected to’* without skill constraints
- Prompt-based edits (for Gemini, Claude, ChatGPT) kept original phrasing intact, while skills showed mixed compliance with explicit rules

## Why it matters

The inconsistency risks miscommunication in professional emails, where hedging (*‘may’* or *‘unconfirmed’*) signals caution. Skills could help but require strict rule-setting; marketers and teams must verify edits manually.

## Sources

1. [When You Ask AI to Fix AI-like Emails, It Adds Unrequested Promises | Finishing Skills That Don't Change the Meaning (Comparing Claude and ChatGPT)](https://note.com/kaisha_ai_kakari/n/n3b128f2eecaf?hl=en) (note.com, 2026-10-12)

Part of the developing story: [AI Tool Showdown From Side Hustle to Kyoto](https://digestai.news/thread/ryo-shows-how-to-use-gemini-chatgpt-and-claude-for-a-20minute-side-hustle) (4 stories)

## Cite

Digest AI, "Claude and ChatGPT add unrequested promises when editing AI-like emails", 12 October 2026, https://digestai.news/story/claude-and-chatgpt-add-unrequested-promises-when-editing-ai-like-email

---

Written by Digest AI's editorial model from the linked sources; the sources are the record. Headlines, digests and key points are written by Digest AI and may be quoted with a link to the story page. Linked articles belong to their publishers. Terms: https://digestai.news/terms#reuse
JSON: https://digestai.news/story/claude-and-chatgpt-add-unrequested-promises-when-editing-ai-like-email.json