August 7th, 2025  – On 2 August 2025, the first set of requirements under the EU AI Act came into effect, introducing stricter obligations for general‑purpose AI (GPAI) providers – including those offering large language models (LLMs).

The new rules require GPAI providers to:

• Transparency – Clearly inform users when content is generated by AI.

• Data Disclosures – Publish a public summary of training data sources and processing methods.

• Risk Mitigation – Assess and reduce systemic risks from powerful AI models.

As Firemind delivers both fine‑tuned models and complete AI systems, we are directly in scope of the EU AI Act. Our Managed AI offering, integrated with the Pulse compliance workflow, embeds transparency, oversight, and governance into every stage of the AI lifecycle. The new regulations highlight the importance of this approach, positioning Firemind and our customers to meet today’s compliance needs and adapt to future requirements.

Firemind’s Compliance-by-design Approach

Managed AI is Firemind’s end‑to‑end AI lifecycle service, embedding compliance, monitoring, and governance into every stage – from use case discovery to post‑deployment oversight. By integrating the Pulse compliance workflow, Firemind ensures transparency, accountability, and readiness for evolving AI regulations.

Our framework addresses compliance across four critical layers:

1. Use Case – Classification and Purpose Clarity
A use case compliance checker workflow powered by a fine‑tuned language model aligned with the EU AI Act guides compliance from the earliest stages of use case discovery and filters out unacceptable or high‑risk scenarios.

2. Data – Quality, Governance, and Bias Control
Continuous and on‑demand evaluation of training, validation, and fine‑tuning data pipelines ensures lawful collection, documentation, version control, and minimisation of bias.

3. Language Model – Robustness and Explainability
Ongoing evaluation of AI models (LLMs, SLMs, open and closed source) ensures they remain secure, resilient, and auditable. Firemind provides detailed documentation for fine‑tuned models to support regulatory and customer transparency needs.

4. Agent/Workflow – Deployment Oversight and Human Intervention
Post‑deployment monitoring and timely human oversight are built into every solution. As an orchestrator, Pulse delivers full logging and documentation of agent logic to enable explainability and traceability.

“The EU AI Act sets a new global standard for responsible AI. We’ve embedded compliance into every layer of our AI solutions so that our customers can innovate confidently and ethically,” said Charlie Hudson, Managing Director at Firemind.

With Managed AI and the Pulse compliance workflow, Firemind ensures that both its own AI systems and those of its customers are secure, compliant, and prepared for future regulatory developments.

About Firemind

Firemind is a generative AI, data, and machine learning consulting partner, trusted by over 200 clients including the Premier League, Vodafone, British Land, and Center Parcs. An AWS Premier Tier Partner with more than 10 AWS Competencies and 16 AWS Service Delivery Programs, we rank among the top AWS partners globally.

Recognised as AWS UK & Ireland Rising Star of the Year 2023, Finalist for AWS Global GenAI Tools Partner 2024, and Top 3 Finalist for NVIDIA Consulting Partner of the Year EMEA 2025, Firemind combines award‑winning expertise with its Pulse generative AI platform and Managed AI services to embed transparency, governance, and compliance into every AI engagement – helping organisations meet and exceed the requirements of the EU AI Act.

Press contacts: Charlie Hudson, Firemind
Email: press@firemind.com
Contact number: +44 020 3468 1670

Get in touch to discover how Firemind can help your organisation stay compliant with the EU AI Act – or learn more and download our Managed AI eBook here.