Live feed · updated daily
Recent policy changes
2965 policy changes
detected across 352 platforms. Most platforms don't announce policy changes — these updates were detected automatically.
August 27, 2026
Removed detailed descriptions of five safety principles (Traceability, Transparency, Agility, Collaboration, Safety by Design) from responsible AI policy; replaced with principle names only.
Why it matters: The updated policy removes explicit, detailed descriptions of how ElevenLabs implements content traceability, maintains transparency about AI interaction, and responds to emerging threats. While the 'Our Safeguards' section remains, users and third-party reviewers no longer have access to the detailed explanations of how these principles operate. This reduces the specificity of ElevenLabs' public safety commitments as stated in its published documentation.
Restructured Enterprise Data Commitments documentation to detail SaaS Platform data retention (30-day automatic deletion), training opt-out controls, and deployment-specific data access rules.
Why it matters: The updated disclosure establishes explicit data retention timelines (30-day automatic deletion) and opt-out mechanisms for training data use on Cohere's SaaS Platform, giving enterprise customers clarity on data handling practices and control over whether their data is used to train Cohere models. Organizations relying on Cohere for regulated workloads or that have made specific data minimization commitments should verify that these disclosed practices align with their vendor contracts and privacy representations to their own customers.
Privacy policy metadata timestamp updated to reflect revised update interval.
Why it matters: This change does not materially affect privacy policy terms or operations. The updated timestamp is a technical metadata adjustment with no bearing on data collection, privacy rights, or user obligations.
Pricing figures updated in privacy policy documentation; no privacy practice changes detected.
Why it matters: The change appears to be a routine administrative update with no apparent impact on privacy practices or user rights. Without broader context about what the figures represent, operational significance cannot be confidently determined.
Price reference updated from $2,457.85 to $2,465.06 in Terms of Service.
Why it matters: The operational significance of this change depends on what metric the updated price represents. Without that context, the practical implications for users, transactions, or platform operations cannot be reliably assessed.
Stay ahead of the changes
You're seeing a fraction of what's changing
ConductAtlas monitors every tracked platform and captures every policy update.
Removed email engagement data collection disclosures; added opt-out mechanism for commercial dataset; narrowed permitted uses of shared customer data.
Why it matters: The removal of email engagement tracking disclosures without replacement language clarifies HubSpot's email data collection practices or reflects a shift in how that practice is documented. The narrowed use-case scope for commercial dataset customers may affect how organizations that rely on enrichment products structure their data workflows. The addition of an explicit removal mechanism provides individuals a new avenue to request deletion of their business contact data from HubSpot's dataset.
Restructured data practices disclosure with new subsections on AI model training, tracking code use, and Intent data sharing, plus clarified data transfer scope.
Why it matters: The updated terms reorganize how HubSpot discloses its AI model training and Website Data practices, making opt-out mechanisms and tracking code use more explicit. This clarifies customer visibility into data handling but does not materially change available protections or opt-out rights; it is primarily a restructuring of existing disclosure language.
Reformatted privacy policy navigation and language selector; no material privacy practice changes.
Why it matters: This change reorganizes the presentation of WhatsApp's privacy policy landing page. Users will encounter updated device platform options and a restructured language selector, but no changes to privacy practices, data handling, or user rights are introduced.
Clarified device platform terminology in Terms of Service footer from OS-specific names to general platform categories.
Why it matters: The updated language reorganizes how WhatsApp describes the platforms on which its service operates, but does not alter the substantive terms governing service access, user conduct, or platform rights. This is a clarification for consistency and does not materially affect operational obligations or user experience.
Acorns privacy policy sentence modified; substantive change not visible in diff
Why it matters: The operational significance of this change cannot be determined because the specific modified sentence is not visible in the provided diff. Compliance teams and users should review the full updated privacy policy directly to understand what may have changed.
Acorns Terms of Service updated; one sentence modified on August 27, 2026.
Why it matters: Acorns is a consumer financial services platform offering brokerage, banking, and investment products. Updates to its Terms of Service may affect dispute resolution, fees, data handling, or user rights. However, the substantive content of this particular change is not visible in the truncated diff provided.
Restructured footer navigation and removed customer support phone numbers from document footer.
Why it matters: The updated Terms of Service footer removes direct phone support contact information and replaces it with a reorganized navigation structure. This is a presentational change only; it does not modify the substantive terms, consumer rights, data practices, or legal obligations established by the agreement.
Removed customer support phone numbers and product listing from privacy notice footer; replaced with simplified navigation.
Why it matters: The updated notice removes customer support contact information and product listings from the footer, which may affect users' ability to locate contact details within the privacy document itself. However, the core substantive disclosures regarding third-party data sharing, categories of personal information, and business purposes remain unchanged, so this reorganization does not alter what data SoFi collects, shares, or processes.
Reorganized Terms of Service footer layout; added product navigation and legal links; removed standalone customer support section.
Why it matters: The updated footer improves navigation to SoFi products, company information, and legal documents without modifying substantive terms. Links to the Privacy Policy and Arbitration Agreement remain accessible, and the change does not alter consumer rights or obligations.
Adds explicit merchant blocking procedure for canceling recurring ACH payments; expands supported card types to include American Express.
Why it matters: The updated terms establish a specific, user-facing mechanism for stopping recurring ACH charges from merchants and clarify the provisional nature of incoming transfers, which aligns operational language with standard banking and ACH processing practices. This gives users a documented procedure for payment cancellation and sets explicit expectations about when transferred funds become available.
Editorial reorganization of Meta's Frontier AI Framework metadata and related article links.
Why it matters: This change does not materially alter Meta's AI governance framework or user obligations. The updates are organizational and editorial in nature.
Updated metadata in GPT-5.5 System Card with new security incident reference
Why it matters: This change provides readers with current security-related documentation and incident analysis rather than outdated blog posts. The metadata refresh maintains the System Card's utility as a reference document without altering the documented safeguards or deployment practices for GPT-5.5.
Added reference to 'The Hugging Face incident and the road ahead' security article in GPT-5 System Card related content
Why it matters: The addition of a linked security article indicates OpenAI is providing supplementary guidance on security incidents and responses. This is informational rather than operational; it does not alter GPT-5's stated safety assessment, capabilities, or user-facing terms.
Added security article reference to Frontier Governance Framework; links to Hugging Face incident analysis dated August 26.
Why it matters: The updated framework makes security incident analysis and response guidance more discoverable within OpenAI's public governance documentation. This improves transparency regarding how OpenAI addresses frontier model safety challenges and security issues.
August 26, 2026
Added 'Brand' link to privacy policy footer navigation; no substantive privacy practice changes detected.
Why it matters: This change does not materially affect OpenRouter's privacy policy. The addition of a 'Brand' navigation link is a formatting update to the footer and does not alter the substantive data handling practices, disclosures, or user rights described in the policy.
Added Brand link to footer navigation; no material change to terms or user rights.
Why it matters: This change does not materially affect how the service operates or what rights and obligations apply. It is a minor update to footer navigation that provides a link to OpenRouter's brand-related resources. No substantive policy, data practice, or user obligation has changed.
Adds consent gate requiring user acceptance of Privacy Policy data practices before service access.
Why it matters: The updated Privacy Policy now gates access to Egnyte services behind an explicit consent mechanism. This affects how users interact with the platform and ensures that use of the service constitutes documented acceptance of Egnyte's data practices. For compliance purposes, the consent gate creates a clear point of user agreement to the Privacy Policy.
Adds explicit consent requirement to Privacy Policy before continued service use.
Why it matters: The updated terms establish an explicit procedural gate requiring affirmative consent before continued service use, creating a clear documentation point for privacy compliance. This affects how users interact with the service and creates a formal consent event that Egnyte can reference for GDPR, CCPA, and similar consent-based privacy frameworks.
Clarified Beta Features definition to include 'preview' designation; explicitly excluded Beta Features from SLA coverage.
Why it matters: The updated terms establish explicit clarity that Beta and preview services carry no service level guarantees or uptime commitments. Organizations relying on experimental features for production use should understand that PlanetScale makes no performance promises for these offerings, which affects continuity planning and resource allocation decisions.
Stay ahead of the changes
You're seeing a fraction of what's changing
ConductAtlas monitors every tracked platform and captures every policy update.
Expanded infringement reporting requirements: now requires prior direct resolution attempt and detailed proof of Checkout.com service use; clarifies Checkout cannot remove third-party content directly
Why it matters: The updated policy establishes a clearer boundary between Checkout.com's role as a payment intermediary and the responsibility of IP rights owners to enforce their rights directly against website operators. The new procedural requirements (direct resolution attempt, service use verification, detailed documentation) may slow infringement report processing but reduce Checkout.com's exposure to claims it failed to act on valid complaints. For merchants, the explicit statement that Checkout.com will not remove content clarifies what they should not expect from the platform, though it does not protect them from payment service suspension if Checkout.com determines infringement has occurred.
Restructured arbitration section with formatting corrections; substantive dispute resolution terms unchanged.
Why it matters: The updated terms do not materially alter Starbucks' dispute resolution policies or consumer obligations. The changes are primarily formatting corrections and structural reorganization of existing language. The substantive provisions governing arbitration, class action waivers, and jury trial waivers continue to apply as previously stated.
Navigation menu updated in privacy policy header; 'Contact' replaced with 'Get a Demo'.
Why it matters: This change does not materially affect privacy policy substance, consumer rights, data handling, or compliance obligations. It is a navigation update only.
Navigation update: 'Book demo' replaced with 'Get Started' in header—no terms of service changes.
Why it matters: This change has no operational significance. It is a website interface update that does not modify the Scale AI Terms of Service, policies, or user obligations. The 'Last Updated' date in the document header remains August 25, 2021, confirming that no substantive terms have changed.
Removed playbook hub reference; replaced with 2026 agentic AI guide announcement on Trust Center.
Why it matters: This change does not materially affect how Writer's platform operates or what users are required to accept. The updated text replaces one marketing callout with another, leaving the core trust and security messaging intact.
Updated promotional content on privacy policy landing page; substantive policy terms remain unchanged.
Why it matters: This change has no material operational significance. It represents a minor refresh to promotional callouts on Writer's privacy policy landing page, while all substantive privacy policy terms, data handling disclosures, and effective dates remain unchanged.
Updated daily. New changes added as detected.