Aleph Alpha announces the launch of PhariaAI: a sovereign-grade enterprise-level artificial intelligence operating system

Aleph Alpha, a German artificial intelligence company, has released PhariaAI, an enterprise-level operating system for generative AI. This new platform combines sovereign design with the interpretability and compliance of large-scale language models.

The platform allows organizations to integrate AI into their core processes while maintaining high transparency and control. PhariaAI aims to meet strict regulatory requirements and includes built-in interpretability and auditability tools, enabling experts to supervise and adjust AI systems to ensure they operate according to organizational standards.

Jonas Andrulis, Founder and CEO of Aleph Alpha, said, "PhariaAI is a game-changer for businesses and governments seeking to leverage AI without compromising control or compliance. Our technology accelerates AI innovation while ensuring transparency, auditability, and comprehensive compliance with organizational standards."

One of PhariaAI's core missions is to simplify the lifecycle of AI, from development to full-scale production. With its range of development tools, including components for fine-tuning, evaluation, and debugging, enterprises and public institutions can quickly scale AI projects using this platform. Additionally, PhariaAI provides ready-made templates for common tasks such as question answering, summarization, translation, and transcription, aiming to accelerate implementation speed and optimize operations.

A key feature of PhariaAI is its hybrid execution capability, which supports flexible allocation of AI workloads in both local and cloud environments. This allows organizations to seamlessly expand their operations while ensuring data privacy and security. This flexibility is crucial for companies handling sensitive information as they can push capacity to the cloud during peak periods while processing confidential data locally.

The core of PhariaAI consists of Aleph Alpha's newly launched Pharia-1-LLM-7B-control and Pharia-1-LLM-7B-control-aligned models, which show significant improvements compared to Aleph Alpha's previous Luminous model series. These models are optimized for key applications and provide adaptability, security, and compliance in various environments. The aligned versions include additional security safeguards to ensure their usability for sensitive conversational AI tasks.

Importantly, both models have been trained to fully comply with GDPR and the upcoming EU AI Act, making them highly suitable for deployment in European businesses and public institutions. These models support seven languages and have been optimized for English, German, French, and Spanish, making them a multilingual AI solution applicable to different markets.

PhariaAI, launched by Aleph Alpha, is an important milestone in the development of enterprise AI in Europe, integrating sovereignty, compliance, and scalability into a powerful platform.