Winner #LatamDigital 2026
Most Innovative AI Solution

Human-directed AI studio producing cinematic content for global brands, with governance, traceability and legal clarity built into every frame.

(Who)ARDE AGENCY S.A.S. — NIT 901.856.077-9, Chamber of Commerce of Medellín reg. No. 0079271812. Studio operating since 2018; incorporated July 19, 2024. Registered address: Cl. 6A # 16-45, Ed. Guayacán de la Calera, Medellín, Colombia.

(What)AI cinematic films and advertising video, hyperreal AI product imagery, hybrid production, branding and brand systems — produced under human direction with generative AI as the engine.

(=Commitments) Deliverables directed & reviewed by humans100% Client data used to train AI models0 Maximum breach notification72h Sub-processor change notice30d

Signable copies available at contact@ardeagency.com.

Standing commitments

ARDE operates a human-directed production model: generative AI is a production tool operated under the direction, curation and final approval of human creative professionals.

• Human review of every deliverableNo AI output reaches a client without direction, curation, editing and approval by ARDE's human team, followed by the client's own approval workflow.
• No training with client dataClient materials, data and deliverables are never used to train general-purpose AI models and are never shared with other clients. Client production runs in enterprise-tier AI workspaces whose agreements contractually exclude customer content from model training (see Higgsfield Enterprise), and on no-training API configurations for other providers.
• End-to-end input traceabilityImages and video are generated exclusively from rights-cleared inputs — the client's real product photography and official brand assets, ARDE-created materials, or licensed assets; never unlicensed third-party works. For every deliverable we maintain a production record documenting which inputs and prompts were used, on which model and platform, and where the human team intervened.
• Likeness & voice consentAny use of a real person's image, likeness or voice requires prior documented consent (model release) explicitly covering AI synthesis. Consent evidence is retained and exhibitable.
• Transparency & labelingWe disclose the use of AI in production (consistent with EU AI Act Art. 50 good practice) and identify AI-generated deliverables or elements so clients can meet platform synthetic-media labeling requirements.
• Prohibited usesNo non-consensual deepfakes or impersonation, no deceptive or harmful misinformation, no unlawful content, no knowing infringement of third-party rights.
• IP clarityDeliverables are assigned to the client on full payment; the jurisdiction-dependent copyright treatment of purely AI-generated elements is disclosed honestly in our Terms (§5).

Security practices

ARDE maintains an internal information-security program covering access control, encryption, secure development, device security, personnel obligations and vendor management. Client materials are treated as restricted information: segregated per project, shared internally on a need-to-know basis, and never reused across clients. A detailed security-practices overview is available under NDA.

Data protection & compliance

In client engagements ARDE typically acts as a processor ("encargado") of any personal data contained in client materials — most commonly the images, likenesses and voices of talent, under documented consent. Our DPA covers instructions-only processing, confidentiality, security measures, sub-processor control with 30-day notice, 72-hour breach notification, audit rights, and return or deletion of data within 30 days of engagement end. We do not sell or share personal data.

ARDE maintains a documented incident-response procedure — detection, triage and severity classification, containment, eradication, recovery and post-mortem — with a commitment to notify affected clients without undue delay and within 72 hours of confirming a personal-data breach. Security reports: contact@ardeagency.com.