The news
A group of 114 members of Congress, led by Rep. Steven Horsford (D-NV) and Sen. Elizabeth Warren (D-MA), sent a letter on October 8, 2026 raising concerns about a proposed deal for Google to buy data from defunct Spirit Airlines to train AI models, The Record reported.
According to The Record, Google proposed paying Spirit $10 million for access to roughly 100 million emails, 500 million Microsoft Teams messages, employment contracts and payroll records. Spirit announced plans to shut down in May 2026, and nearly 1,000 employees lost jobs in the Las Vegas area after the closure, the report said.
The lawmakers questioned whether removing identifying details would protect former employees. Their letter said that stripping names, email addresses or other direct identifiers does not necessarily make a dataset anonymous. They urged safeguards including independent confidentiality reviews, limits on how the data may be used and input from employees on how deidentification is done, according to The Record.
A Google spokesperson told The Record the company is not looking to buy any personal information from Spirit, and that the information will either be excluded entirely or deidentified by an independent third party before Google receives it.
The same day in the United Kingdom, the Information Commissioner's Office (ICO), the country's data-protection regulator, said it had secured commitments from 10 AI developers: Amazon, Anthropic, Apple, Cohere, DeepSeek, Google, Meta, Microsoft, OpenAI and Stability AI, The Register reported. The companies pledged clearer explanations of how personal data is used in training, better ways for people to exercise their data rights and stricter assessments of safeguards. xAI was dropped from the group for separate investigation, according to the report.
The ICO also acknowledged that current training practices still raise legal compliance problems around sensitive personal data, and opened a six-week call for evidence on agentic AI, with responses due November 20, 2026, The Register said. Richard Nevinson, the ICO's director of technology regulation, said AI's benefits depend on trust and transparency.
The numbers
- Lawmakers signing letter
- 114
- Proposed Google payment to Spirit
- $10 million
- Emails / Teams messages in scope
- About 100M / 500M
- AI firms making ICO pledges
- 10
- ICO agentic AI evidence deadline
- November 20, 2026
Why CEOs should care
For CEOs and CFOs, the Spirit case shows that internal records are now a saleable asset, and that selling them can trigger political scrutiny. Emails, chat logs and payroll files are created under expectations of confidentiality by employees and customers. Before licensing any such data for AI training, including in a restructuring or asset sale, boards should ask: who consented, what do employment contracts and privacy notices allow, and would deidentification survive a determined re-identification attempt?
General counsel and privacy officers should treat deidentification claims skeptically, as the lawmakers did. Require an independent review, written limits on downstream use and a right to audit the buyer. Bankruptcy trustees and acquirers should note that data sales can draw congressional attention even when legal.
For companies buying AI services, the ICO pledges give procurement a checklist. Ask vendors how they explain personal data use in training, how data-subject requests are handled and what safeguards assessments they run. The 10 named developers have now made those commitments to a regulator, so asking them to show it in contracts is reasonable.
The bigger picture
AI developers are hunting for fresh, high-quality text as public web sources tighten access, and private corporate archives are an obvious target. Regulators on both sides of the Atlantic are signaling that this market will be watched, with US lawmakers focused on deal terms and the UK regulator on ongoing practices.
What’s next
Watch whether Google or Spirit's estate responds to the letter and whether the deal proceeds, and for the ICO's findings from its agentic AI call for evidence, which closes November 20, 2026.
What “Fact-checked” means
Fact-checking means testing a story’s facts against the evidence before it is published. This story went through at least two separate checks before this version was published.
- What we checked
- Its names, figures, dates, job titles, quotes and who said what were checked against the story’s sources, including its main source where it could be opened. The headline was checked for accuracy and overstatement.
- How
- A first check reviewed the whole story. If it passed, a second, skeptical check went back to the sources to look for mistakes in the most important facts. If a check flagged the story, it was edited to fix the problems found, and a separate re-check then reviewed the whole story again.
- Who
- The checks are made by our newsroom, as steps kept separate from the writing, under rules set by our editor, Hussein Mukhtar. A story the checks still flag is held for the editor, who decides whether it is fixed, published or dropped.
- If something is wrong
- “Fact-checked” does not mean error-free. If a material error is found after publication, we correct the story and add a note saying what changed. Report an error
Companies in this story
