Updated October 2026
The Enron emails: the corporate archive that came before the data market
The Enron email dataset is the best-known corporate mail archive in public research. It holds about 0.5M messages from about 150 users, most of them senior managers, and it reached the web because the Federal Energy Regulatory Commission posted it during its investigation of the company.1
The corpus
What the Enron collection contains
The collection is narrow by today's standards: one company, a leadership-heavy set of about 150 mailboxes, and roughly half a million messages. The version researchers work from now is dated 2015-05-07. In our reading, its appeal was simple: real business correspondence, in volume, that anyone could read.1
We make no claim here about which AI models were trained on it; we have no sourced fact that settles that. What is documented is how it became public: a regulator released it during an investigation, not through a sale by the company.1
| Fact | Value and source |
|---|---|
| The Enron email corpus | About 0.5M messages from about 150 users, mostly senior management, posted to the web by the Federal Energy Regulatory Commission during its investigation; the current research version is dated 2015-05-07 |
Two decades on
Set beside Spirit Airlines' archive
Spirit's bankruptcy filings describe about 100M emails across 80,000 accounts, plus 500M Teams items and employee records going back to August 1986. Where Enron's archive was a few hundred thousand messages from a management circle, Spirit's spans an entire workforce and several tools. Both came from companies in collapse: Enron's surfaced through a federal investigation, Spirit's through a Chapter 11 sale process run by the debtor's advisers.2
The route out is different too. Spirit's data went to a court auction where Google became the successful bidder at $10.0M. As of 2026-10-14 the court had yet to approve the sale, and the hearing is listed for 2026-11-06. The archive was also to be de-identified, with Tonic.ai certifying each transfer to the court under the terms the ombudsman recommended.3,4,5
The paid market
From public record to licensed asset
Smaller companies now sell mail through wind-down channels. SimpleClosure reports 150+ data deals for wound-down startups, typically $10K-$100K each, and says email sits in scope alongside Slack, code and docs, with sensitive employee and health records excluded.6,7
Cielo24 is a reported example: 13 years of Slack, Jira and email, sold for what was described as "hundreds of thousands of dollars". Enron's messages, by contrast, reached the public through a regulator rather than a buyer.8,1
For sellers
What the Enron story suggests to a seller today
Our reading: Enron shows how long a corporate archive stays in circulation once it is out. The research version still in use is dated 2015-05-07. A company licensing its mail today sets the terms first: who receives it, in what form, and after what scrubbing. SimpleClosure, for one, says it processes each dataset for PII before licensing.1,9
There is a second contrast. Enron's mail came out whole, with names attached. Spirit's buyers were judged partly on how they would strip identities: a declaration filed after that auction said de-identification and privacy factors may be outcome-determinative. The expectation that a seller scrubs the archive first is newer than the Enron corpus.10
FAQ
Questions, answered with sources
About 0.5M messages from about 150 users, mostly senior management.1
The Federal Energy Regulatory Commission posted them to the web during its investigation.1
The current research version is dated 2015-05-07.1
We do not answer legal questions. Ask the licensing company: their team reviews this with you before anything is signed.
Before you sign
Legal questions go to the licensing company
We do not answer legal questions. Ask the licensing company before you sign: their team reviews these with you.
Questions for this kind of business
- If our archive includes mail from former employees, how do CCPA, GDPR and state privacy laws apply to licensing it?
- How are union members' messages and NLRA questions handled in an email license?
- How do CCPA, GDPR, UK GDPR and state privacy laws apply, and is de-identified data still personal information?
- How are employee data, union contracts and NLRA questions handled?
Questions for any buyer
- Is the license exclusive, and for how long?
- Can you sublicense the data, and to whom?
- Which AI labs will receive it, and will you tell us?
- How long do you keep it, and how is it deleted?
- What is your liability cap, and who indemnifies whom?
- Will the contract ban re-identification of people in the data?
- Who removes personal data, by what method, and do we approve it first?