1. You are interacting with an AI system (Art. 50(1))
Nexus Legal is an artificial intelligence system. In compliance with Article 50(1) of Regulation (EU) 2024/1689 (the EU Artificial Intelligence Act), we inform you clearly and from the first moment of use that the analyses, searches and drafts produced by the platform are generated by AI systems.
This information is additionally displayed inside the product through the "AI System" badge visible in the working interface, so that no user can reasonably be unaware of interacting with a machine.
System provider: Quantum Nexus Ventures FZCO (Dubai, United Arab Emirates), with an EU representative designated under Article 27 GDPR (see the Privacy Policy).
2. Intended purpose
Nexus Legal is designed to assist legal professionals (lawyers, advisors, in-house legal teams) with document analysis, case law and legislation research, and the preparation of working drafts.
It does not replace the advice of a qualified, licensed professional. The system's outputs are decision-support tools and always require qualified human review before any use with legal effects.
The system is not intended to be used by judicial authorities to research or interpret facts or the law, to apply the law to a concrete set of facts, or for alternative dispute resolution with legal effects (the scenario in Annex III, point 8, of the Regulation).
3. Capabilities
Multi-agent analysis: a primary analysis node (Node A), an adversarial audit node that challenges and validates the conclusions (Node B) and, optionally, a risk and red-team node that attacks the user's own position (Node C).
Deterministic citation verification: references to statutes and rulings produced by the system are automatically and deterministically checked against official corpora (BOE, CENDOJ and equivalent sources in other jurisdictions), flagging verified and unverified citations.
Case law and current legislation search: semantic search over legislation and case law corpora across multiple jurisdictions, with source attribution.
4. Known limitations
Like any system based on language models, Nexus Legal can make mistakes: inaccuracies, incorrect citations, debatable interpretations or omissions of relevant statutes or rulings.
Deterministic citation verification reduces the risk of non-existent or erroneous citations, but it does not guarantee the complete absence of error nor the exhaustiveness of the analysis.
For this reason, human oversight is mandatory: every output must be reviewed by a qualified professional before use. The system is not suitable for use by a judicial authority to decide cases, nor to replace human professional judgement.
5. Human oversight and control measures
Although the system's classification does not require the measures of Article 14 of the Regulation (reserved for high-risk systems), Nexus Legal voluntarily incorporates oversight mechanisms aligned with that standard:
Cross-audit (Node B): every primary analysis is subjected to an adversarial audit by a second, independent agent that looks for weaknesses, contradictions and unsupported claims.
Confidence index (ISO 31000): reports include a confidence index structured according to the ISO 31000 risk management methodology, so professionals can calibrate the weight they give to each conclusion.
Human review state and review flags: every output carries a review state (pending / reviewed / approved) and flags marking the points that require priority attention from the human reviewer.
6. AI-generated content (Art. 50(2))
System exports (reports, dossiers, draft pleadings) are marked as AI-generated content, through a visible indication in the document and machine-readable metadata (including a traceability log with an integrity hash, the model chain used and the human review state).
This in-product marking currently applies within the scope of the Spain jurisdiction, where the first rollout of the Regulation's transparency framework is active. This transparency page is of general scope and available to all users.
7. Data and models
Nexus Legal uses a chain of third-party language models accessed via API — currently Anthropic (Claude), DeepSeek and Google (Gemini), with automatic failover between them; firms may also configure their own provider (BYO), in which case the document goes only to the provider they chose. Quantum Nexus Ventures FZCO does not train models on client documents.
PII Gatekeeper: before any content is sent to the model providers, an anonymisation layer replaces direct personal identifiers with opaque codes. This layer reduces the risk of personal data exposure, although it cannot guarantee the complete elimination of residual data.
Zero Retention: in Analysis Mode, client documents are processed entirely in memory and are not persisted in database or on disk. Details of processing activities, retention modes and GDPR safeguards (including the Article 27 EU representative) are set out in the Privacy Policy and the Data Processing Agreement (DPA).
8. Governance and complaints
In Spain, the supervisory authority for artificial intelligence is the Spanish Agency for the Supervision of Artificial Intelligence (AESIA).
For complaints related to the operation of the AI system you can write to support@nexusquantum.legal; we will handle and document every complaint. This does not limit your right to address the competent authorities.
For personal data protection matters, the competent authority in Spain is the Spanish Data Protection Agency (AEPD): www.aepd.es.
9. Regulatory framework and classification
Regulation (EU) 2024/1689 of the European Parliament and of the Council of 13 June 2024 (Artificial Intelligence Act). According to our assessment, Nexus Legal is a limited-risk AI system, subject to the transparency obligations of Article 50, and not a high-risk system under Annex III: it is not intended for use by judicial authorities (Annex III, point 8) but by private legal professionals as a support tool.
Spain — Organic Law Bill 121/000096 (in parliamentary passage), which implements the Regulation at national level: designation of AESIA as supervisory authority, the sanctioning regime and complaint channels. We monitor its passage and will adapt this page to its final text.
This classification is reviewed periodically and whenever there is a relevant change in the system's functionality or in the regulatory framework.
10. Contact
For any question about this transparency statement or about the operation of the AI system:
Provider: Quantum Nexus Ventures FZCO · Dubai, United Arab Emirates.