Gensyn Challenges Big Tech with the Industry's First Auditable AI Model

Open-1b lets anyone independently verify how it was trained, setting a new standard for transparency and accountability for future AI systems.

LONDON, Sept. 15, 2026 /PRNewswire/ -- Gensyn, the industry's leading decentralized AI lab, announced the launch of open-1b, a verifiable AI model designed to allow anyone to test and independently verify how it was trained.

As AI plays a growing role in human life, influencing personal and business decisions, transparency into its development remains closed off. Users have little insight into how these systems are built, trained, or shaped by data. Meanwhile, leaders from the world's largest AI companies continue to warn about the risks of the technology's rapid advancement.

Ben Fielding, Co-Founder and CEO of Gensyn, explained, "Auditable AI means a model's decisions do not have to come from a black box like Anthropic or OpenAI models. Today, Gensyn has unveiled the toolkit that allows someone to inspect the data that shaped it, verify the training process step by step, and understand why it reached a particular conclusion. To achieve our goals with Gensyn, we need AI models that anyone can trust and verify themselves without relying on a company's version of the truth. Open-1b is a tremendous milestone for the AI industry, proving a path toward verifiable AI, an absolute necessity for the future of intelligence."

How open-1b compares to existing AI models:

• Closed AI models (e.g., Claude, ChatGPT): Provide access to model outputs, while keeping training methodologies entirely proprietary.
• Open-weight models (e.g., Llama): Distribute model weights, but omit the underlying training process.
• Open-recipe models (e.g., OLMo): Offer both the model weights and the instructions used for training.
• Auditable AI (Gensyn's open-1b): Delivers the model, the training recipe, and verifiable proof allowing third parties to independently audit any stage of training.

Simply sharing an AI model's code and datasets is insufficient for true independent verification, as hardware variations frequently lead to inconsistent results. Gensyn addresses this fundamental challenge by providing infrastructure that ensures training outcomes can be accurately reproduced and verified across diverse hardware. Demonstrating this capability, open-1b lays the groundwork for fully auditable artificial intelligence.

Gensyn is moving AI transparency beyond disclosure to proof, giving people the tools to verify how a model was trained for themselves.

About Gensyn
Gensyn is a decentralized AI lab building forecasting systems that continuously improve by learning from real-world outcomes. Its technology combines verifiable AI infrastructure, which allows predictions to be checked against independently verified outcomes, with incentives for people to contribute new information that models have not previously seen. Together, these capabilities create a continuous cycle of prediction, verification, and learning.

Gensyn is backed by a16z crypto, CoinFund, Galaxy Digital, Eden Block, Maven 11, and more. For more information, visit gensyn.ai.

Media Contact
Michelle O'Connor
‪(415) 578-0156
[email protected]

SOURCE Gensyn