MITRE Federal AI Sandbox to assist with AI experimentation and prototyping capability for US government agencies
Non-profit organization MITRE is building a new capability intended to give its artificial intelligence (AI) researchers and developers access to a massive increase in computing power. The new capability, MITRE Federal AI Sandbox, will provide better experimentation of next-generation AI-enabled applications for the federal government.
The Federal AI Sandbox is expected to be operational by year’s end and will be powered by an NVIDIA DGX SuperPOD that enables accelerated infrastructure scale and performance for AI enterprise work and machine learning.
As U.S. government agencies seek to apply AI across their operations, few have adequate access to supercomputers and the deep expertise required to operate the technology and test potential applications on secure infrastructure.
“The recent executive order on AI encourages federal agencies to reduce barriers for AI adoptions, but agencies often lack the computing environment necessary for experimentation and prototyping,” says Charles Clancy, MITRE, senior vice president and chief technology officer. “Our new Federal AI Sandbox will help level the playing field, making the high-quality compute power needed to train and test custom AI solutions available to any agency.”
MITRE will apply the Federal AI Sandbox to its work for federal agencies in areas including national security, healthcare, transportation, and climate. Agencies can gain access to the benefits of the Federal AI Sandbox through existing contracts with any of the six federally funded research and development centers MITRE operates.
Sandbox capabilities offer computing power to train AI applications for government use including large language models (LLMs) and other generative AI tools.
It can also be used to train multimodal perception systems that can understand and process information from multiple types of data at once such as images, audio, text, radar, and environmental or medical sensors, and reinforcement learning decision aids that learn by trial and error to help humans make better decisions.
“MITRE’s purchase of a DGX SuperPOD to assist the federal government in its development of AI initiatives will turbocharge the U.S. federal government’s efforts to leverage the power of AI,” says Anthony Robbins, vice president of public sector, NVIDIA. “AI has enormous potential to improve government services for citizens and solve big challenges, like transportation and cyber security.”
The NVIDIA DGX SuperPOD powering the sandbox is capable of an exaFLOP of performance to train and deploy custom LLMs and other AI solutions at scale.
In March, MITRE launched its AI Assurance and Discovery Lab to discover and mitigate critical risks in AI-enabled systems that need to operate in increasingly complex, uncertain, and high-stakes environments. MITRE was joined by Sen. Mark Warner, Rep. Donald Beyer, and Rep. Gerry Connolly as it opened its lab.
Related
