Fujitsu Kozuchi AI technologies assist AKOS AI in delivering solutions for EU AI compliance

• 22 Apr 2025
• TM Team
• 0 Comments

Fujitsu Limited and AKOS AI announced that Fujitsu has provided the Italy-based startup with AI Trust technologies, consisting of five core tools from its Fujitsu Kozuchi AI service. These technologies will enable AKOS’s AI governance platform AKOS HUB, to offer EU AI Act compliance, risk management and general AI governance services and solutions to enterprise customers.

AKOS AI offers AKOS HUB, which provides a set of packages of no-code/low-code tools and AI governance, supporting organizations of all sizes and industries in ensuring fairness, robustness, and risk management. AKOS HUB is comprised of the following packages that leverage Fujitsu Kozuchi :

1. AI Act Package: Simplifies EU AI Act compliance

This package is based on the AI Compliance Assistant. By answering a questionnaire users determine their AI system's risk level and receive a list of required actions. The system then creates a compliance report. AKOS HUB also offers packages for ISO 42001 compliance the international standard for AI management systems, and the AI Risk Management Framework (AI RMF) of the U.S. National Institute of Standards and Technology (NIST).

2. AI Impact Assessment Package: Identifies risks based on ethical guidelines

This package is based on the AI Ethics Impact Assessment. It evaluates AI systems by mapping components, identifying stakeholder interactions, and automatically pinpointing ethical risks. A comprehensive risk analysis table, based on EU guidelines and past incidents, is generated.

3. AI Risk Assessment Package : Ensures the reliable and efficient implementation of AI system risk management

This package is based on the AI Risk Management tool, which supports trustworthy AI development throughout its lifecycle, from planning to operation. It facilitates ethical risk identification, bias mitigation, and robust tracking of evidence for problem resolution and auditing. It ensures compliance with AI regulations (e.g., EU AI Act) and provides powerful visualization of AI projects for proactive governance.

4. AI Fairness Package : Verifies and improves the fairness of AI models

This package is based on the Fujitsu AI Ethics for Fairness and mitigates AI bias across all phases using Fujitsu’s intersectional bias detection algorithm, ensuring fairness across multiple attributes (e.g., gender, age, ethnicity). It evaluates fairness and prediction accuracy, clarifying trade-offs for optimal model selection.

5. LLM Package : Diagnose biases hidden within large language models from diverse perspectives

This package is based on the LLM Bias Diagnosis and evaluates biases across LLMs on various ethical topics. Users can assess publicly available LLMs or their own LLM. Bias evaluation covers four Sustainable Development Goals areas: Climate Action, Gender Equality, Healthcare, and Education. It also provides services for evaluating LLMs based on OWASP security guidelines.

Surya Josyula, Head of Business Incubation Division, Technology Strategy Unit, Fujitsu Limited, comments:

“As an active contributor to several leading organizations including the Organization for Economic Cooperation and Development (OECD), and the Global Partnership on Artificial Intelligence (GPAI), and as a recent pledge to the Hiroshima AI Process International Code of Conduct, Fujitsu has taken the lead in collaborating closely with partners, standardization institutes and industry regulatory bodies dedicated to setting global standards for AI trust and ethics. Fujitsu is a technology pioneer, and our goal is to research, develop and deliver new solutions based on these standards and promote safe, secure and trustworthy AI for society”.

Alberto De Lazzari, AKOS AI Chief Executive Officer and Co-Founder comments:

“The increasing adoption of AI solutions presents companies with new challenges, such as the ethical and fair use of AI, which require sound governance practices. AKOS HUB, the AI Governance Platform, supports responsible AI by helping enterprises manage and oversee the legal, ethical, and operational aspects of AI.”

The European Union AI Act, which came into force on August 1st, 2024, is the first comprehensive legal framework aimed at regulating AI within the EU. The Act classifies AI systems based on their risk levels, and its primary objectives are to ensure that AI systems safeguard individuals' health, safety, and fundamental rights, while addressing the risks posed by AI models.

Future Plans

Fujitsu and AKOS AI recognize AI's potential as a catalyst for enterprise transformation and sustainable growth. The two companies’ vision is that AI is not only beneficial for people but also for the planet, ensuring a harmonious balance between innovation and environmental stewardship.

Fujitsu and AKOS AI are also committed to developing secure AI systems to ensure the prevention of threats and attacks that could compromise the confidentiality, integrity, or availability of an AI model or system powered by AI.