UMNAI: Revolutionizing AI Compliance in the Face of New EU Regulations

The rapidly evolving landscape of Artificial Intelligence (AI) has been a topic of intense discussion and scrutiny, particularly in the context of regulatory compliance. A recent study by Stanford University has brought to light that most AI models, including those developed by tech giants like Google and OpenAI, do not comply with the forthcoming EU AI Act. This Act, which has received substantial support in the European Parliament, aims to regulate AI usage and mitigate its potential risks.

The Stanford study assessed 10 major AI model providers against 12 of the 22 requirements stipulated in the Act. The requirements were categorized into four areas:

• data resources,
• compute resources,
• the model itself, and
• deployment practices

Significant discrepancies in compliance levels were found, with some providers scoring below 25%. There was also a considerable lack of transparency among model providers.

Several areas of non-compliance were highlighted, including the failure to disclose the status of copyrighted training data and undisclosed energy usage and emissions during model training. The study also noted disparities between open and closed AI model providers.

The study concluded that all the examined foundation models fall short of compliance with the current regulations outlined in the AI Act draft. It also underscored the challenges that the high-level obligations established in the AI Act may pose for many companies.

In response to the Act, executives from 150 prominent companies expressed their concerns about the stringent regulations in an open letter, warning that these limitations may prompt companies to consider leaving the EU and investors to withdraw their support for AI development in Europe.

The study suggests there is an urgent need for enhanced collaboration between policymakers and model providers in the EU to effectively address the gaps and challenges, and find a common ground to ensure the appropriate implementation and effectiveness of the AI Act.

Hybrid Intelligence?is a?unique and transformative approach to AI?that has emerged from over four years of intense deep-tech development by UMNAI, a European start-up.?

Hybrid Intelligence, seamlessly merges advanced AI with human insights and control?natively?enhancingpredictive accuracy, transparency, and regulatory compliance.?

UMNAI's Hybrid Intelligence Framework incorporates the XNN (eXplainable Neural Nets) Platform an easy-to-use developer platform that generates predictive models in a unique and powerful neuro-symbolic architecture.

XNN models are inherently transparent, completely predictable, auditable, fully explainable and provide best-of-breed predictive performance. This changes the nature of risk and compliance management as these powerful AI models are as traceable, auditable, and predictable as traditional expert systems.

XNNs?provide for a clear understanding of?precisely?how?a decision?is made, which is a key requirement of the EU AI Act.

The Hybrid Intelligence Framework introduces robust and certain risk and compliance management.

Hybrid Intelligence, at the foundation of any AI tech stack, drives higher efficiency, increased profitability and verifiable compliance.

The Stanford study underlines the urgent need for enhanced collaboration between policymakers and model providers in the EU. UMNAI's Hybrid Intelligence Framework offers a practical and effective solution to the challenges identified in the study.

As we move towards a future where AI plays an increasingly significant role in business and society, technologies like UMNAI's Hybrid Intelligence must emerge to prioritise transparency, compliance, and the responsible use of AI.

UMNAI is raising a Seed Round - investors are welcome.