MANAGEMENT UPDATE.

A SANDBOX FOR AI

On January 30, a new initiative was announced: The creation of The AI Learning and Innovation Hub Sandbox (AI Hub). It’s been designed to support government agencies in testing and scaling AI solutions to improve workforce development economic mobility, and public benefits access. According to a release, “The AI Hub provides a secure environment where AI tools, such as job placement prediction models, can be tested using public or de-identified data.”

Overall, according to the release, “the AI Hub aims to empower government agencies to make informed decisions and implement AI solutions effectively and responsibly, ultimately improving service delivery and outcomes for residents.”

This initiative is being fostered by Social Finance, a non-profit that works with the public and private sectors to build innovative partnerships and investments and HumanServices.ai, an AI consultant.

Amber Ivey, a Vice President of Impact Advisory at Social Finance, shared her enthusiasm for the AI Hub’s potential impact with us: “Government agencies often lack a structured way to assess AI tools,” she said. “This is about more than just testing technology—it's about building the knowledge base and skill set government leaders need to make informed, responsible choices about AI.”

Stephen Rockwell, Founder of HumanServices.ai, added “Our government, philanthropic, and industry partners see the value of implementing technology and processes for the development of ethical and effective AI models and tools. In such a fast-evolving domain, the AI Learning and Innovation Lab is a vital technology and technical assistance partner to government agencies and executive offices learning and applying AI to real world use cases.”

The Sandbox provides a safe, and no-cost, environment in which agencies can look at a variety of AI solutions for specific challenges in order to determine how effective, useful and secure each one is in real-world scenarios. “By testing various solutions,” according to the release, “agencies can make informed decisions and choose the best fit for their needs—ensuring they select a tool that has been tested for their specific use case and dataset before procurement.”

Additionally, the hub has developed guidelines and promising practices to help address some of the outstanding challenges in the world of AI, including data privacy and algorithmic bias.

In an effort to extend its reach, the AI Hub intends to share models and datasets through a centralized platform. The effort here is to give other agencies, who haven’t necessarily engaged in the process, to learn from others and gather background about the ways various solutions can work for them.

Finally, the AI Hub is offering training – including workshops and hands-on experiences -- to help provide AI literacy and technical skills for government personnel.