Live feed · updated daily
Recent policy changes
3515 policy changes
detected across 352 platforms. Most platforms don't announce policy changes — these updates were detected automatically.
July 29, 2026
Navigation reorganization in updated Terms and Conditions; no material substantive changes detected.
Why it matters: This change does not materially affect how the service operates or what users agree to. It represents document reorganization rather than policy substantive changes.
DeepL privacy policy revision detected; substantive impact unclear from available change context.
Why it matters: The detected change appears to be a minor formatting or navigation adjustment rather than a substantive modification to DeepL's privacy practices or user data handling. Without visibility into the full text of the modified sentence or the surrounding policy context, the operational significance cannot be reliably determined.
Reordered supported features list for Mistral Medium 3.5 model card; no capability or pricing changes.
Why it matters: This change has minimal operational significance. The reordering of feature documentation does not alter which capabilities are available, how they function, or under what terms they are offered. Developers will encounter the same API endpoints and features; only their position in the published documentation has changed.
Added reference to Cohere Trust Center in Acceptable Use Policy documentation resources.
Why it matters: The updated documentation now directs users to the Cohere Trust Center for additional governance and trust-related resources. This is primarily a navigation enhancement rather than a substantive policy change, and does not materially alter operative terms or user obligations.
Formatting update in Privacy Policy price display; no substantive privacy practice changes.
Why it matters: The change does not materially affect privacy practices, consumer rights, or operational terms. This is a display adjustment only.
Stay ahead of the changes
You're seeing a fraction of what's changing
ConductAtlas monitors every tracked platform and captures every policy update.
Modified a numerical value in Terms of Service from $1,946.33 to $1,922.45
Why it matters: A numerical value was modified in OpenSea's binding terms; without visibility into the figure's purpose, operational significance cannot be reliably determined. Organizations using the platform should review the full updated document to confirm whether this change affects pricing, fees, thresholds, or other material terms.
Removed cookie consent interface and accept/manage buttons from privacy policy landing page; users directed to separate cookie policy.
Why it matters: This change affects how users access cookie preference controls when landing on Canva's privacy policy. The underlying cookie consent mechanisms are not eliminated, but users can no longer manage cookie choices directly from the main privacy policy page; they must navigate to a separate cookies policy instead.
Updated mailing address for customer service requests from San Francisco to Cleveland across billing and consent procedures.
Why it matters: The updated address ensures that customer written correspondence reaches Affirm's correct location. Customers submitting billing disputes, consent withdrawal requests, or other written inquiries must use the new Cleveland address to ensure timely processing and response.
Refreshed related news links on Frontier AI Framework documentation; no substantive policy changes.
Why it matters: This change does not materially affect the Frontier AI Framework itself. It is a documentation maintenance update that refreshes supporting news links without modifying any substantive AI governance, model risk, or control language.
Updated dispute resolution to favor consumers' local courts and required 30-day notice before most Terms changes, while establishing new Automated Data Collection compliance obligations.
Why it matters: The updated terms materially alter dispute resolution for consumers by replacing universal California jurisdiction with a consumer-favorable approach where consumers can use their home country's courts and laws. This change affects how complaints are filed and resolved. Additionally, the introduction of separate Automated Data Collection Terms with revocable permission creates new compliance obligations for users employing bots or crawlers, and the 30-day advance notice requirement provides greater transparency into future Terms modifications.
Technical update to OpenAI Usage Policies page formatting; substantive policy language unchanged.
Why it matters: The detected change is technical rather than substantive. No modification to operative policy language, enforcement provisions, restrictions, or user obligations occurred. Users continue to operate under the same stated usage policies.
July 28, 2026
Restructured User Content storage terms into three categories with defined access limitations and separate retention documentation
Why it matters: The updated terms establish a clearer distinction between when OpenRouter stores content for debugging/review (opt-in), for processing a specific request (temporary), and when it may be retained longer (per feature documentation). This reorganization defines the scope and purpose of storage more explicitly and adds a contractual restriction stating OpenRouter will not access processing-related storage except for operation, security, troubleshooting, or legal compliance. For users sending batch or large-volume requests, this clarifies both the temporary nature of storage and the limits on how that content may be used.
Updated privacy request contact email addresses; most requests now route to support@sourcegraph.com instead of privacy@sourcegraph.com
Why it matters: The updated email addresses determine how quickly and efficiently users and organizations can exercise data subject rights under CCPA, state privacy laws, and other regulations. Consolidating contact points may reduce routing delays, but users must use the correct address to ensure their request reaches the appropriate team.
Privacy policy landing page updated with product promotions; underlying privacy terms unchanged.
Why it matters: This change does not materially affect privacy protections or data handling. It is a product marketing update to the policy page header promoting Modal's Kimi K3 product and API offering. The substantive privacy policy remains dated May 2023.
Product announcement added to Terms header: Kimi K3 launch and Shared API with token-based pricing model.
Why it matters: This change announces new Modal products and pricing in the Terms header but does not modify the contractual framework, user obligations, or rights. The operational significance is minimal as the substantive terms remain unchanged.
Added regional language and geographic market navigation to Privacy Policy header
Why it matters: This change improves user accessibility by offering the Privacy Policy in multiple regional languages and adding navigation links for specific geographic markets. No substantive privacy practices, data protections, or user rights were modified.
Stay ahead of the changes
You're seeing a fraction of what's changing
ConductAtlas monitors every tracked platform and captures every policy update.
Language menu reorganized in Terms of Use header. No substantive policy changes.
Why it matters: This change has no material operational significance. It is a reorganization of the page navigation header that does not affect the substantive terms, consumer rights, data practices, or policy obligations.
Added Thai language version to privacy policy regional selector.
Why it matters: The change extends Samsung's privacy policy accessibility to Thai-speaking users by providing a localized language option. This is an operational and accessibility improvement with no changes to substantive privacy terms, rights, or obligations.
Removed supplementary license documentation and disclaimers from DeepSeek-R1 MIT License; core terms unchanged.
Why it matters: The MIT License text itself remains unchanged, so the legal rights and obligations governing use of DeepSeek-R1 software are unaffected. The removed explanatory text and disclaimers were supplementary; their absence does not modify the core license permissions or conditions.
Reformatted privacy policy header with updated branding and navigation structure.
Why it matters: This change updates the document header and branding structure but does not modify any substantive privacy terms, data collection practices, or consumer protections. The operational effect is visual and navigational only.
Timestamp updated in Replit Terms of Service footer; no policy changes detected.
Why it matters: This change has no operational significance. A timestamp in the document footer was adjusted, but no Replit policies, user obligations, data practices, or platform rights were modified.
Timestamp correction in privacy policy footer; no operational changes to privacy practices or terms.
Why it matters: This change has no material operational significance. The timestamp adjustment does not modify privacy practices, data handling, user rights, or platform obligations.
Clarified AI model training opt-out conditions in Data Processing Addendum; simplified from dual-condition opt-out structure to single customer opt-out requirement.
Why it matters: The updated language clarifies a key operational mechanism for customer control over AI training use of their data. By consolidating the opt-out condition from a dual-requirement structure to a single customer opt-out, the revised DPA reduces ambiguity about when Mistral AI may proceed with model training and provides customers with a clear, simplified path to prevent such use. This is particularly material for organizations subject to GDPR or equivalent frameworks, where lawful basis documentation and customer transparency regarding automated processing are regulatory requirements.
Removed assurance that paid API and Enterprise Input/Output data exclude from AI model training; expanded Input definition to include third-party integration data.
Why it matters: The removal of explicit training exclusions for paid services may affect vendor contracts and Data Processing Addenda that organizations relied on to satisfy their own privacy commitments. The expanded definition of Input data to include third-party integration data means Mistral now processes a broader scope of user information for its stated business purposes. Together, these changes may require organizations to renegotiate vendor agreements and revise customer-facing privacy disclosures to reflect the updated scope of data collection and processing.
Restructured help documentation; introduced gas sponsorship guidance.
Why it matters: This change reflects a documentation update to Uniswap's help resources. The addition of gas sponsorship guidance signals the platform's intention to provide user-facing information about this feature, though the change itself does not establish new operational rights or obligations for users.
Extends error-reporting deadline from 60 to 120 days for non-Send Money transactions.
Why it matters: The extended 120-day error-reporting window gives account holders significantly more time to identify transaction discrepancies before the dispute window closes. Under Wise's prior terms, an error discovered after 60 days would be outside the reporting period; under the revised terms, the same error reported within 120 days can be escalated to Wise Support.
Modified one sentence in Privacy Policy; specific operational change unclear from available diff context.
Why it matters: Privacy policy changes affect how user data is collected, used, and disclosed. Without visibility into the specific sentence modification, the operational significance cannot be determined. Compliance teams should obtain the full text to evaluate regulatory applicability and any required updates to vendor agreements, privacy notices, or internal policies.
Single sentence modified in OpenSea Terms of Service; pricing display updated from $1,872.77 to $1,946.33.
Why it matters: The operational significance of this change cannot be determined from the provided diff context alone. The modification involved a single sentence with an associated price figure change, but without the complete text of the modified language, the practical implications for users or the platform remain unclear.
Adds protection for mobile messaging opt-in data, expands data transfer mechanisms, updates EU/UK representative contacts, enables phone submissions for data access requests.
Why it matters: The updated statement establishes explicit protections for messaging campaign consent data that were not previously stated in explicit language, reducing potential third-party data sharing and aligning Zoom's practice with GDPR consent principles. The expanded data transfer frameworks provide operational flexibility for Zoom's international operations while the updated representative contacts and new phone channel for data requests improve accessibility for EU, UK, and Swiss data subjects seeking to exercise their privacy rights.
Removed redundant section headers from 1% match promotion terms; substantive offer terms and fee structures unchanged.
Why it matters: This change has no material operational significance. The 1% match offer terms, eligibility criteria, fee structures, and holding requirements remain identical. The revision is a formatting optimization that removes duplicate headers without altering substantive policy.
Updated daily. New changes added as detected.