The productive artificial intelligence tools, whose name we hear more and more every day, had taken over the entire technology world with its development that surprised us all. However, there were great concerns about the various risks these models posed, from harmful content to data security.
An important move in this regard came from technology giants. Leading in the field of artificial intelligence Microsoft, OpenAI, Google and Anthrhopic; He founded a group called “Frontier Model Forum”.
The group will aim to ensure the safe and responsible development of productive AI.
The group of four companies wants to ensure that productive AI models develop safely and responsibly. The models that will be prioritized will be “large machine learning models capable of performing a variety of tasks and exceeding the capabilities available in the most advanced models available.”
In the blog posts shared by the companies, it was stated that the new security group will have four main goals. These targets were listed as follows:
- Enhance AI security research to encourage responsible development of models, minimize risks, and provide independent, standardized assessments of capabilities and security.
- Identifying best practices for responsible development and deployment of models by helping people understand the nature, capabilities, limitations and impact of technology.
- Collaborating with policymakers, academics, civil society and companies to share information about security risks.
- Support efforts to develop applications that can help tackle society’s biggest challenges, such as climate change mitigation, early detection and prevention of cancer, and tackling cyber threats.
The Frontier Model Forum will begin its work by establishing an advisory board that will offer advice on artificial intelligence and share information with companies and governments. Let’s also mention that the group wants to bring the world of technology together. In the statement, it is stated that its doors are open to other companies that are trying to take steps in artificial intelligence security.
34430
