← All pages

1. What the document is: Meta's Acceptable Use Policy for the Llama API — a prohibited-use list applying to "Services as defined in the Llama API Terms."

What this wiki found — complete, every page

Presence facts, not verdicts — each flagged term links to its definition and the exact clause on this page. Absence of a badge means the term isn't currently flagged here, not that the page is risk-free.

Llama API – Acceptable Use Policy

Dataset: GenGA (Generative AI Governance Archive) — 11 AI services, 2025–present

GenGA (Generative AI Governance Archive) is this wiki’s Generative-AI dataset: raw policy snapshots captured by the Open Terms Archive genai-eu project across 11 GenAI providers (ChatGPT, Claude.ai, DeepSeek, Google Generative AI Services, Le Chat, Llama API, Meta AI, Microsoft Copilot, Perplexity, Qwen Chat, xAI). Unlike PGAv2, GenGA has no pre-tagged risk scores — all risk scoring and keyword tagging on these pages is LLM-assigned by direct reading, spanning 2025–present.

Note: This page contains documented policy clauses. Risk assessment is qualitative and context-dependent. For analysis of patterns across platforms, see:

• AI Act Substance Matrix
• Tier Bifurcation
• Scraper Enforcement Gap
• Methodology

Source: sources/GenGA/Llama API/Acceptable Use Policy/ (raw Markdown captures, Open Terms Archive genai-eu dataset — no pre-tagged JSONL exists for this source) Last updated: 2025-11-13 (content unchanged since; see methodology note)

Note on methodology: No pre-computed risk_score/keyword fields exist for this source. Keyword tags and clause analysis on this page are LLM-assigned. Treat as first-pass analysis, not externally verified ground truth. This wiki’s risk-scoring system was retired project-wide on 2026-06-21 (see methodology.md §4) — no page, GenGA included, computes or displays a numeric risk score.

Content-hash dedup: This doc type has 84 raw captures (2025-11-11 through 2026-04-15) but only 4 distinct content-versions by SHA-256 hash. One hash (41 occurrences, alternating throughout the entire date range) is a transient scrape error page (“This page isn’t available — An unexpected error has occurred”), confirming the live page intermittently failed to render for the scraper roughly every other check — these are excluded below as failed captures, not real policy states. The remaining 43 captures collapse to one substantive content version (first seen 2025-11-13) plus two cosmetic variants missing only the page’s top breadcrumb-link header (2026-01-24, 2026-04-15) — the policy body itself never changed across the entire dataset window.

Overview

1. What the document is: Meta’s Acceptable Use Policy for the Llama API — a prohibited-use list applying to “Services as defined in the Llama API Terms.”

2. Input rights / 3. Output restrictions: Not addressed in licensing terms; only conduct restrictions on what Outputs may be used to generate or claim.

4. Non-explicit predatory clauses: None — but two developer-facing obligations are notable: (a) a ban on “Representing that the use of AI API Services outputs are human generated” (an anti-deception rule, narrower than a blanket disclosure mandate, same structure as Google’s equivalent clause); (b) an affirmative duty to “appropriately disclose to end users any known dangers of your AI system” — a genuine AI-disclosure-style obligation imposed on developers building atop the Llama API.

5. Regulatory references: None found — no GDPR or EU AI Act citation.

6. Regional variation: Not addressed.

7. Key risk to users: Low — a standard prohibited-use list (violence, CSAM, illegal weapons/drugs, fraud, defamation, spam, deceptive human-generated-content claims) plus the two developer-disclosure obligations noted above.

Flagged Keywords & Risks (LLM-assigned)

• ai disclosure — “Fail to appropriately disclose to end users any known dangers of your AI system” is a prohibited use (i.e., developers must disclose known AI-system dangers); separately, “Representing that the use of AI API Services outputs are human generated” is banned. Why it matters: a developer-facing (B2B) AI-disclosure obligation, distinct from the consumer-facing disclosure clauses tracked elsewhere in this dataset.

Regulatory & Research Context

Llama API was not among the providers directly studied by Davidson et al. (2026) or Pandit et al. (2026) in this batch, but applying Pandit et al.’s general finding by analogy — that GenAI terms place “responsibilities [users] cannot materially fulfil without violating the terms” — the duty flagged above to “appropriately disclose to end users any known dangers of your AI system” pushes a similarly open-ended compliance burden onto third-party developers rather than Meta itself. Edwards et al. (2025) frame this kind of liability-shifting as characteristic of the “platformisation paradigm,” in which the provider positions itself as a neutral intermediary while developers absorb the practical disclosure obligations.

Changes Summary

Version History

2025-11-13

• Explicit AI clause: YES
• Non-explicit predatory: NO
• Flagged keywords: ai disclosure

Clause: ai disclosure

3.5. Representing that the use of AI API Services outputs are human generated [is prohibited]. 4. Fail to appropriately disclose to end users any known dangers of your AI system [is prohibited].

All other captures in this dataset (excluding the 41 transient-error-page repeats) are byte-identical to this version through 2026-04-15.