Get the weekly research letter
Companies change their terms quietly. We read every version and catch what actually changed. One email a week on the changes that matter and what they mean. No account.
The terms prohibit developers from using platform-sourced data to make decisions about or to target users based on sensitive personal characteristics including health status, financial information, race, ethnicity, political or religious beliefs, sexual orientation or gender identity, or union membership.
This analysis describes what Meta's agreement states, permits, or reserves. It does not constitute a legal determination about enforceability. Regulatory applicability and practical outcomes may vary by jurisdiction, enforcement context, and individual circumstances. Read our methodology
This provision establishes categorical prohibitions on specific uses of platform data that intersect with anti-discrimination law and data protection frameworks governing special categories of personal data, creating compliance obligations for any developer whose application processes or could infer such attributes from platform data.
The updated terms remove developer restrictions that previously applied to large-scale platforms, meaning companies with more than 700 million monthly active users can now access Meta's AI models without special permission. The terms also eliminate the restriction on EU developers accessing multimodal models and consolidate account management under either managed accounts or Meta accounts. Developers previously subject to these restrictions should review whether the updated terms now permit their use case.
View change record →The updated terms authorize Meta to retain user-submitted content if its systems flag the content for a potential policy violation, in addition to retention tied to legal compliance and contractual rights. This expands the circumstances under which content may be preserved without explicit time limits. Under the revised language, content retention decisions may now be driven by automated policy-violation flagging in addition to legal or contractual necessity. Developers integrating the Llama API should understand that flagged content may be retained indefinitely pending policy review.
View change record →This new provision prohibits discriminatory uses of platform data across protected categories, addressing algorithmic bias and civil rights concerns in AI/ML applications.
View full change record →Under this clause, developers are prohibited from using data obtained through Meta's platform to make decisions about or target users based on health, financial status, race, ethnicity, political or religious beliefs, sexual orientation, gender identity, or union membership, regardless of whether such data was explicitly provided by the user or inferred.
How other platforms handle this
In certain circumstances, the right to data portability, which means that you can request that we provide certain Personal Data we hold about you in a machine-readable format
If you want to see what information we have collected about you, you can request a copy of your data in the Data & Privacy section of your User Settings. You should receive your data packet within 30 days.
For data portability requests, We will select a format to provide Your personal information that is readily useable and should allow You to transmit the information from one entity to another entity without hindrance.
Monitoring
Meta has changed this document before.
Receive same-day alerts, structured change summaries, and monitoring for up to 20 platforms.
"Don't use Platform Data to make decisions about the following sensitive categories or use Platform Data to target people based on sensitive categories including health, financial information, race, ethnicity, political or religious beliefs, sexual orientation or gender identity, or union membership.Excerpt from Meta's Llama API Terms of Service
1. REGULATORY LANDSCAPE: This provision directly engages GDPR Article 9 (special categories of personal data), which restricts processing of health, racial, ethnic, political, religious, trade union, and sexual orientation data. It also intersects with US anti-discrimination law, the Equal Credit Opportunity Act (ECOA) for financial targeting, the Fair Housing Act for housing-related decisions, and the FTC Act for deceptive or discriminatory advertising practices. The Irish DPC, EU national supervisory authorities, and the FTC are primary enforcement bodies. 2. GOVERNANCE EXPOSURE: High. The prohibition encompasses both explicit use of such data and inferential use, which creates compliance complexity for machine learning applications that may derive sensitive attribute inferences from platform data without explicitly processing those categories. Developers in advertising technology, financial services, insurance, or employment should conduct a specific assessment. 3. JURISDICTION FLAGS: EU/EEA developers face the highest exposure under GDPR Article 9's strict special category regime. Illinois developers may face additional exposure under BIPA if biometric data is involved. Financial services developers face FTC, CFPB, and banking regulator interest in discriminatory targeting practices. Healthcare developers face HIPAA and FTC Health Breach Notification Rule implications. 4. CONTRACT AND VENDOR IMPLICATIONS: Advertising technology vendors and marketing platforms receiving platform data from developers must be contractually prohibited from using that data for sensitive category targeting. Standard DSP and ad network agreements should be reviewed to confirm compliance with these restrictions. 5. COMPLIANCE CONSIDERATIONS: Compliance teams should audit all algorithmic systems that process platform data for any inferred or explicit use of sensitive category attributes, update model governance documentation, review advertising targeting configurations to exclude prohibited categories, and implement technical controls that prevent sensitive category use in automated decision systems. GDPR Data Protection Impact Assessments should be conducted for any processing that involves or could involve special category data.
Regulatory citations, enforcement risk, and due diligence action items.
Ad personalization controls removed. Contact scanning added. Advertiser data partnerships quietly dropped. A timeline of every change.
Provision-level monitoring, governance timelines, and regulatory mapping built from archived source documents and historical version tracking.
This provision establishes categorical prohibitions on specific uses of platform data that intersect with anti-discrimination law and data protection frameworks governing special categories of personal data, creating compliance obligations for any developer whose application processes or could infer such attributes from platform data.
Under this clause, developers are prohibited from using data obtained through Meta's platform to make decisions about or target users based on health, financial status, race, ethnicity, political or religious beliefs, sexual orientation, gender identity, or union membership, regardless of whether such data was explicitly provided by the user or inferred.
ConductAtlas has identified this type of provision across 295 platforms. See the full comparison.
No. ConductAtlas is an independent monitoring service. We are not affiliated with, endorsed by, or sponsored by Meta.