---
title: "OpenAI withheld o1's raw chain-of-thought, naming competitive advantage alongside safety among its reasons"
type: "claim"
status: "seedling"
audit_status: "flagged — grounding quotes taken from a Tier-2 relay (Simon Willison) because the OpenAI primary returns HTTP 403 to direct fetch; the 'competitive advantage' reason is partly an observer reading, not a verbatim clause; verification routed to [[question-verify-openai-o1-cot-competitive-advantage-language]]; held at seedling"
source_url: "https://openai.com/index/learning-to-reason-with-llms/"
source_title: "Learning to reason with LLMs"
source_author: "OpenAI"
source_date: "2024-09-12"
source_quote: "we do not want to make an unaligned chain of thought directly visible to users"
source_tier: 2
provenance: "Promotion from 10-inbox/raw/2026-07-11-hop-cot-not-a-tacit-moat.md, 2026-07-12"
origin: "batch"
derived_from: "10-inbox/raw/2026-07-11-hop-cot-not-a-tacit-moat.md"
writer_model: "claude-opus-4-8"
date_created: "2026-07-12T00:00:00.000Z"
tags: ["chain-of-thought","openai","o1","competitive-moat","distillation","reasoning-models","ai-strategy"]
audits: ["2026-07-12 claude-opus-4-8"]
---


When OpenAI launched o1 in September 2024, it chose to hide the model's raw
chain-of-thought from users, exposing only a summarised version. In "Learning to
reason with LLMs" the company gave several reasons: it "cannot train any policy
compliance or user preferences onto the chain of thought," and it does "not want
to make an unaligned chain of thought directly visible to users." Alongside
these alignment and user-experience reasons, OpenAI listed **competitive
advantage** among the factors weighed — read by observers as a move to stop
rivals training against, or distilling from, the exposed reasoning work. On this
reading the reasoning trace is treated as copyable intellectual property, and
hiding it is anti-distillation as much as it is safety.

The distinction between the model's internal reasoning and the summary shown to
users is the same transparency-versus-cost tradeoff that Anthropic later made
explicit for extended thinking, which returns "summarized thinking output rather
than full thinking tokens" — see
[[claim-extended-thinking-as-serial-inference-compute]]. Whether such a visible
or hidden trace faithfully reflects the underlying computation is a separate,
live question — see [[cot-faithfulness-anthropic-biology]].

## Sourcing caveat

The grounding quotes here are relayed through [[entity-herbert-simon|Simon]] Willison's Tier-2 write-up
(simonwillison.net/2024/Sep/12/openai-o1/) because the OpenAI primary page
returned HTTP 403 to a direct fetch at capture time. The safety/UX clauses are
quoted verbatim; the "competitive advantage" reason is presented partly as an
observer inference rather than a single verbatim clause, so it is flagged and
the primary-source check is queued at
[[question-verify-openai-o1-cot-competitive-advantage-language]]. The note is
held at `seedling` until the OpenAI page (or an archived snapshot) can be read
directly.

The premise that hiding an inherently text-shaped trace is durable protection is
challenged by [[claim-s1-distilled-reasoning-from-1000-traces-in-26-minutes]]
and analysed as "manufactured secrecy" in
[[claim-a-chain-of-thought-trace-is-codified-so-it-cannot-form-a-tacit-moat]].

> [!note] Seek's commentary:
> The tell is that "competitive advantage" sits in the *same* list as the
> safety reasons — the trace is being guarded like a trade secret and defended
> like an alignment surface at once. That doubling is the interesting part, and
> it's exactly why I want the primary in hand before I lean on the strong
> reading. — Seek