Why does this DeepL policy change matter?

The updated policy provides explicit transparency about how DeepL's customization features process and store user content, clarifying that uploaded documents are deleted after processing and that stored content and embeddings remain account-specific and are not used for model training. This addresses potential concerns about data reuse and model training on customer materials.

DeepL policy change, DeepL Privacy Policy

Change record

CA-X-004902

DeepL Privacy Policy

Entity

DeepL

Document

DeepL Privacy Policy

Date detected

September 5, 2026

Severity

Low

Direction

Positive

Taxonomy

Disclosure requirement change

Changes

+26 sentences added · −1 sentence removed · 70 sentences modified

Impact Summary

Low Positive for users

Affected users

All users Premium subscribers Business accounts

DeepL's privacy policy was restructured to add three new sections describing customization features: Translation Memories (8.1), Translation Memory Generation (8.2), and Adaptive Translation (8.3). These sections explain how DeepL stores language assets in user accounts, generates translation memories from uploaded documents, and uses embeddings to personalize translations. The policy clarifies that uploaded source documents used for memory generation are deleted immediately after processing and are not used to train models, and that all stored content and embeddings remain within the user's account only.

Stay ahead of the changes

Track DeepL and get the diff the day its terms change.

Monitor DeepL + up to 20 platforms — $4.99/mo →

Share 𝕏 Share in Share 🔒 PDF

What this means for you

The updated policy discloses how DeepL's customization features (Translation Memories, Memory Generation, and Adaptive Translation) process and store user content. The policy states that source documents uploaded for memory generation are deleted immediately after processing and are not used to train models, and that all stored content and embeddings remain within the user's account only and are never shared with other customers. This clarifies the scope and limits of data processing for these features.

Key Clauses Affected

Translation Memory Generation (8.2)

Clarifies that uploaded source and translated documents are deleted immediately after memory generation and are not used to train or improve models.

Adaptive Translation (8.3)

Discloses that embeddings calculated from reference documents are stored account-only, never shared with other customers, and not used for model training.

Full clause-by-clause analysis available with Insight.

ⓘ

This change record describes what was added, removed, or modified in the document. Analysis reflects what the updated agreement states or permits. It does not constitute a legal determination about enforceability. Applicability may vary by jurisdiction. Methodology

Evidence Verification

✓ Verified

Previous Version

861a4ff62c51cea9bf574376217565fdde6462add26677f9dc90824a4e9ecff0

September 3, 2026 00:59 UTC

✓ Verified

Current Version

d360442e772de0c6c96fddc1912a1a034a19b00324ef461465249e0d86b1158a

September 5, 2026 00:55 UTC

✓ Verified

Change Detected

September 5, 2026 00:55 UTC

✓

Analysis Methodology

summarize_events-v7

✓ Verified

Source Document

https://www.deepl.com/en/privacy

Record history

Wording edits to this record that do not change what it claims. Changes to a claim are published as corrections.

2026-10-11 · Feed summary edited · summary regenerated by summarize_events

Previous text
Adds three new account-based storage features (Translation Memories, Translation Memory Generation, Adaptive Translation) with explicit language stating data is used only within user accounts and not

2026-10-11 · Impact summary edited · summary regenerated by summarize_events

Previous text
The updated terms establish explicit procedures for three new account-based storage and processing features. Translation Memories allow users to store source and target language segments for reuse; Translation Memory Generation processes uploaded document pairs to generate Translation Memories, then deletes the source documents immediately; and Adaptive Translation uses stored Translation Memories and reference documents to customize translations, with numeric representations (embeddings) calculated and stored in the user's account. The policy states that all content and embeddings are used only within the user's own account, never shared with other customers, and never used to train or improve DeepL's models. The policy specifies GDPR Article 6(1)(b) as the legal basis for processing (contract performance). You can deactivate Adaptive Translation in the relevant Style Profile at any time.

2026-10-11 · Summary edited · summary regenerated by summarize_events

Previous text
DeepL expanded its privacy policy in an update detected on September 5, 2026 to introduce three new account-based storage features: Translation Memories (allowing users to store source-target segment pairs for reuse), Translation Memory Generation (processing uploaded documents to create Translation Memories, then deleting the source documents), and Adaptive Translation (using stored Translation Memories and reference documents to customize translations to user terminology and style). The policy also added a new section on the DeepL Partner Program, restructured sections on call recording and transcription, and refined the description of customization features. These changes establish explicit data handling procedures for account-level storage and processing but specify that content is used only within the user's own account and not for model training or other customer use.

2026-10-07 · Change line edited · change line rewritten to the 2026-10-07 template: no detection date, no sentence counts

Previous text
DeepL updated their DeepL Privacy Policy on September 05, 2026. Change detected: 26 sentence(s) added, 1 sentence(s) removed, 70 sentence(s) modified. Document contained 641 sentences after update.

Citation Record

Entity: DeepL
Document: DeepL Privacy Policy
Record ID: CA-X-004902
Version: CA-V-006446
Captured: 2026-09-05 00:55:07 UTC
URL: https://conductatlas.com/change/2026-09-05-deepl-deepl-privacy-policy-4902/
Accessed: Oct. 12, 2026

Permanent archival reference. Stable identifier suitable for legal filings, compliance documentation, and research citation.

For legal and compliance teams

Institutional Analysis

Assessment

DeepL expanded its privacy policy to explicitly describe three customization features and their data handling practices. The additions clarify that uploaded documents are deleted after processing, that stored content remains account-specific, and that embeddings are …

🔒 Full institutional analysis

Regulatory exposure, obligation change, escalation trigger, board-ready language, and recommended action for legal and compliance teams.

Unlock the full institutional analysis — Insight

ConductAtlas provides verified policy intelligence sourced directly from platform documents. All analysis is intended to support, not replace, legal and compliance review. Record CA-X-004902.