Corrections log

Hugging Face updated its Terms of Service on March 19, 2026 to replace references to 'Enterprise Support' with 'Expert Support' across two locations. The first change modified the list of offerings that Services may include, and the second change renamed the support tier itself. This is a terminology update that affects how the support offering is named and labeled in the agreement, with no stated change to the scope or features of the support tier itself.

Stability AI updated its Terms of Service on March 19, 2026 to replace references to 'Brand Studio' with 'Dream Studio' throughout three sentences. The change affects the scope of services covered by the terms and the specific provisions governing user-created model adaptations. This appears to be a product naming or rebranding update with no material change to the underlying rights, obligations, or restrictions stated in the agreement.

Twilio modified its Terms of Service on March 19, 2026, removing Brazil from its entity-specific contracting structure while adding Japan as a separate jurisdiction with its own Twilio entity. The agreement also revised the definition of Order Form to remove a clause permitting self-service online purchases, narrowed the scope of service modifications by stating Twilio will not 'materially decrease' overall functionality, and made formatting corrections to 'Third Party Services' terminology. The net effect is a restructuring of which Twilio legal entity customers contract with based on domicile, and a contractual commitment that service changes will preserve overall functionality.

Headspace restructured its privacy policy on March 19, 2026, removing the detailed table of contents and adding navigation links to related privacy documents including a Consumer Health Data Privacy Policy and HIPAA Notice of Privacy Practices. The previous version organized privacy disclosures under ten numbered sections (Collection, Use, Sharing, Security, Rights, Children's Privacy, Cookies, Changes, and Contact). The updated version removes this structural organization, integrating related policies through links instead of a single comprehensive document.

Nintendo's privacy policy was updated on March 19, 2026 with several revisions to language describing data collection, retention practices, and third-party certifications. The policy now states it collects error information from both users and devices (previously only devices), removed specific examples of location data uses like checking into Nintendo locations, simplified retention language to reference only 'applicable law' without detailing sensitivity-based practices, eliminated detailed disclosure of persistent identifiers collected from child users, and changed its privacy certification from CARU (Children's Advertising Review Unit) to ESRB (Entertainment Software Rating Board). These changes alter how Nintendo describes its data practices and child data handling, though the operational scope of collection and use remains largely similar.

Klarna removed structural formatting elements ('Show more' expandable sections) from its Privacy Policy on March 19, 2026, and reorganized how it presents data collection methods. The policy previously listed data collection in a numbered format with 'Personal data that you provide to us' as item 1 followed by a description; the updated version integrates this description directly into the opening sentence without the numbered structure. This is a formatting and organizational change with no material alteration to the substantive privacy practices described.

Klarna reorganized the table of contents and navigation structure of its Terms of Service on March 19, 2026. The updated document reorganizes the listing and grouping of legal agreements, adding explicit headers and categories such as 'Legal agreements for Klarna services', 'Pay later', 'Money Transmissions', 'Monthly Financing', and 'Memberships'. The document now cross-references additional service agreements including the WebBank Klarna Credit Card Agreement and reorganizes how users navigate to payment plan terms, financing options, and related disclosures. This is a structural and organizational change to how the terms are presented rather than a substantive modification to user rights or obligations.

Cash App reorganized its Terms of Service by moving the 'Generative AI Terms of Use' section from section XXIII to section XXII, and renumbering the 'Other Legal Terms' section from XXIV to XXIII accordingly. This is a structural reorganization with no material change to the substance of the terms themselves. All cross-references throughout the document were updated to reflect the new section numbers.

FanDuel's privacy policy was updated on April 18, 2026 to reflect two document corrections. The policy's last-update timestamp was changed from April 28, 2026 to April 28, 2025, and the policy scope was expanded to explicitly include FanDuel's Fantasy Picks website (www.fanduel.com/picks) and its related mobile app as covered services subject to the privacy policy. Previously, these services were not explicitly listed in the policy's scope definition.

SoFi substantially revised the cookie and tracking technology section of its Privacy Notice on April 21, 2026. The prior language stated that SoFi shares tracking information with social media, advertising, and analytics partners, and that non-selection constitutes agreement to these technologies. The updated language introduces a Privacy Preference Center, describes cookie categories (Functional, Performance, Strictly Necessary), explains that blocking cookies may impact site functionality, and provides explicit mechanism to manage consent preferences by category rather than opting into all tracking by default.

Robinhood expanded the scope of its privacy statement on March 13, 2026 to explicitly cover Robinhood Social, a social media product, and reorganized how it describes regulatory applicability across its financial and non-financial services. The updated policy states that financial products remain subject to the Gramm-Leach-Bliley Act, while Robinhood Social is subject to state privacy laws. The practical difference is that users of the social media product now have explicit notice that different privacy rules apply to that service compared to financial products.

Coinbase modified formatting and navigation elements in its User Agreement on March 11, 2026. The substantive indemnification language remained unchanged. The document now displays interactive section headers with visual toggle indicators (󰅒 symbols) for the Individual User Agreement and Coinbase Business User Agreement sections, along with a Share button for the document. These are formatting and navigation changes with no impact on the contractual terms users operate under.

Okta's Privacy Policy was updated on May 5, 2026 to add a trailing space at the end of a sentence about opt-out mechanisms for third-party vendor cookies and device identifiers. The sentence previously read 'Alternatively, you can opt out of a third-party vendor's use of cookies by visiting the Network Advertising Initiative opt-out page or control the use of device identifiers by using their device's settings.' and now ends with a space character. This is a formatting change with no operational impact on the terms or consumer rights.

OpenAI's privacy policy underwent substantial restructuring in an update detected on March 10, 2026. The policy removed multiple standalone sections including 'Disclosure of Personal Data', 'Retention', 'Data controls', 'Your rights', 'Children', 'Security', 'Additional U.S. state disclosures', and 'Changes to the privacy policy', consolidating or relocating their content. The policy added explicit links to jurisdiction-specific versions for EU/EEA/UK/Switzerland and Korea residents, and modified language describing how training data is collected from publicly available internet sources, removing prior references to advertiser data partnerships and replacing them with broader sourcing disclosures.

Google Gemini added one line to its Gemini Apps Privacy Notice on April 17, 2026: 'Manage or delete the Gemini app on your Android device'. This appears in the Related Resources section at the end of the document. The addition provides a direct reference to app management and deletion instructions for Android users, but does not modify any privacy practices, data collection disclosures, or rights.

YouTube Ads updated the language selection in their Terms of Service on March 8, 2026. The document previously offered terms in English and Vietnamese (VN), and now offers them in English and Thai (TH). This is a localization change that makes the terms accessible to Thai-speaking users, though it does not alter the substantive content of the agreement itself.

PayPal removed two sections from its Purchase Protection Program policy on March 6, 2026: a navigation menu listing specific claim types (QR Code and In-Person Transactions, Item Not Received Claims, Not as Described Claims, Ineligible Items, Transaction Eligibility, Online Dispute Resolution, Opening Disputes timeframes, and Dispute options) and a reference to 'Government Relations' in the footer. The core eligibility language and program terms remain unchanged. This appears to be a structural or navigation simplification with no change to the substantive protection terms themselves.

Netflix's updated privacy statement on March 6, 2026 removed specific US state privacy disclosures from the main document and consolidated regional privacy information. The statement simplified language around data collection (changing 'register with' to 'create your account', 'Offers' to 'promotional offers') and removed explicit mentions of voice inputs and inferences about household ad preferences from the usage and advertising information sections. The statement now describes advertising data collection more narrowly, omitting the prior disclosure that Netflix makes inferences about ad preferences for you or your household.

TikTok restructured its Community Guidelines on March 6, 2026, reorganizing content policies into a new categorized format with expanded explanatory sections. The previous version presented guidelines as general commitments to safety and integrity with links to external resources. The updated version explicitly lists policy categories (Youth Safety, Hate Speech, Misinformation, etc.) and states that each category includes 'More information' sections with key term definitions, common questions, and examples. The operational change is organizational and disclosure-focused: users and reviewers now have a single consolidated policy document with explicit navigation structure rather than distributed guidance across multiple pages.

Microsoft updated three discrete elements in its Responsible AI Principles document on March 6, 2026. The introductory call-to-action statement changed from 'Accelerate business growth with trustworthy AI' to 'Build your business with trustworthy AI', and the supporting descriptor was revised to emphasize 'safe and secure AI business practices' rather than risk reduction. A resource reference changed from 'Watch the webinar' to 'Get the e-book'. Finally, language about Copilot security requirements was altered from 'all your existing security and compliance requirements are inherited' to 'all your existing security and compliance requirements are inherited', with minor grammatical adjustments ('When using Copilot' to 'When you're using Copilots'). These are primarily formatting, presentational, and grammatical refinements with no material change to stated policy or obligations.

Microsoft updated its data retention policy on March 6, 2026, to provide more specific guidance on how long it keeps your data and under what circumstances. The new language clarifies that retention depends on factors like whether you expect Microsoft to keep data until you delete it, whether automated deletion controls exist, and the sensitivity of the information. It also explains what happens after you delete items, including a 30-day window before permanent removal from their systems.

Microsoft removed a sentence from its privacy statement that described consent-based marketing contact via auto-dialer and prerecorded voice technology potentially generated using AI. The updated document no longer explicitly discloses this practice. The removal creates a gap between what the policy previously stated users might consent to and what it now describes about marketing contact methods.

DeepL updated its Privacy Policy homepage on April 29, 2026 by replacing promotional content in the header section. The previous version featured event promotion language; the updated version now directs users to a Spring Event Hub and references infrastructure expansion announcements. These are navigational and promotional changes to the policy landing page, not substantive modifications to privacy protections, data collection practices, or user rights.

DeepL's Terms and Conditions document underwent a minor update on May 5, 2026, where a single sentence in the introductory or promotional section was modified. The change replaced marketing language about 'This is a test' with a new phrase about 'The hidden cost of caption churn in multilingual meetings.' This appears to be a content refresh rather than a substantive change to user rights, obligations, or service terms. The core legal provisions governing your use of DeepL services remain unchanged.

we may share your contact information and other content provided (e.g., name, email, phone number) and history of home or other property searches, including those you may have saved or liked, with a real estate professional