Our approach to AI
At Nymbl, we use AI to enhance and accelerate our VFX workflows, not to replace human craft. We focus on AI-assisted enhancement rather than unguided generation, ensuring every project remains firmly artist-led and creatively driven.
This policy outlines how we use AI in ways that are transparent, responsible and aligned with your expectations as a client. It reflects how we work with you, how we protect your data, and how we make decisions about AI.
Transparency and tools
- Assistive, not generative: AI improves efficiency and visual fidelity under the strict direction of our VFX artists.
- Our toolkit: We use a combination of frontier models via secure commercial frameworks and open-source models on locally hosted infrastructure.
- Client disclosure: We remain transparent whenever AI workflows play a substantial role in project delivery.
You can always ask us how AI was used on a specific deliverable: we will give you a straight answer.
Security and IP protection
Protecting pre-release assets and client Intellectual Property (IP) is our top priority. We maintain rigorous security controls and certified cybersecurity standards across our studio.
- Absolute IP control: For projects requiring strict confidentiality, we deploy AI workflows entirely on local, isolated infrastructure. Client plates and project files are never exposed to public models or third-party servers.
- Secure commercial frameworks: When using external frontier models, we strictly use enterprise APIs that guarantee zero data retention for model training.
- Data protection: We operate in full compliance with the UK Data Protection Act 2018 (UK GDPR).
Compliance and copyright
Our AI practices align with the laws of England and Wales.
- Ownership: Because our AI use is heavily directed by human artists, the resulting work remains proprietary. We ensure all assets are fully cleared and ownable, actively avoiding third-party copyright infringement.
- Regulation: We adhere to the UK’s pro-innovation AI framework, prioritising accountability and safety.
- International standards: While governed by UK law, we support global clients by monitoring and adhering to international frameworks, including the EU AI Act and emerging US regulations.
Addressing AI bias and ethics
We recognise that AI models can inherit visual or cultural biases. We mitigate this through strict human oversight and by fine-tuning our local models, ensuring all AI-supported visuals are accurate, inclusive and representative. We use AI to support, not substitute, creativity.
Our governance model
We apply simple but effective oversight:
- All client-facing outputs are checked by a human.
- One team member acts as our internal AI lead.
- We review AI use and this policy quarterly.
This approach ensures consistency, compliance and creative integrity.
Responsible AI use
Nymbl does not use AI to mislead, deceive or infringe upon the rights of others. We regularly audit our local nodes, APIs and overall cybersecurity posture to ensure we maintain the highest ethical and technical standards.
Policy review
We will update this policy quarterly or sooner if legislation or best practice changes. If you ever want more detail, or to discuss how we use AI on your projects, just ask.
Questions
For questions regarding our AI workflows or security infrastructure, please contact hello@nymbl.com.