Vinay Singh, who is fresh off a two-year stint as the CFO and chief artificial intelligence officer (CAIO) at the U.S. Department of Housing and Urban Development (HUD), said the agency needs to stay focused on literacy and innovation in its development of AI technologies to benefit operations and mission delivery.
In a recent interview with MeriTalk, Singh talked about the challenges, progress, and potential of AI in the workforce at HUD, and said increasing access to the technology and awareness of its uses is critical to AI’s integration within the agency.
“The number one thing, without a doubt, was AI literacy,” Singh said. “We focused on defining generative artificial intelligence [and] how it can impact the mission – whether it’s protecting the mission or amplifying the mission,” he added.
Singh said HUD’s internal AI Governance Board (AI@HUD) worked well and was critical to the success of developing policies and processes for AI use throughout the department because it collaborated across programs and support offices.
“The governance aspect, in our executive committee meetings, our other committee meetings, were functioning very well, including discussions with the [Office of Inspector General]” Singh said.
Singh highlighted HUD’s collaboration with the private sector to increase AI literacy and innovation throughout the department and develop better tech applications as a result. He said the agency hosted master classes and provided AI trainings for all staff to help introduce innovative solutions with private sector partners to better articulate HUD’s mission and goals with AI – all in the interest of bettering citizen outcomes and experiences.
“Of course, industry was in a fast-paced sales mode,” Singh said. “But for HUD to build a sustainable foundation around AI and prove to HUD that the private sector is aligned to the mission, we needed to engage them more,” he added. “I would advise industry that you have to come in with some smarts and some awareness of what our [HUD] focus and priorities are,” Singh recommended.
Singh admitted that an area where the AI@HUD team struggled was making progress with union partners due to fears that AI implementation would end up leading to job losses.
“Our AI@HUD Executive Committee was nearly unanimous in saying, ‘We need to continue to engage with the unions, we’re making progress, let’s negotiate,’” Singh said. “It was just held that it was too sensitive an area. Not everyone wanted to open that door,” he added. Unfortunately, historical relations for decades within HUD around technology and unions stifled progress. “We were successful to initiate meaningful dialogue in the first several months given this was not just any old tech, but mission-changing technology – but the fact that we couldn’t, I couldn’t, articulate why this was different,” Singh added, “is my biggest regret.”
Singh reassured that AI implementation into work tasks is not likely to lead to people losing their jobs. “It’s not about cutting jobs, it’s about adding balance,” Singh said. “The message to the union is, ‘don’t give up’, come to the table with tangible asks like they’ve done and just keep moving forward,” he added.
Singh said he sees AI’s usefulness in many areas of HUD operations including document summarization, training and performance evaluation, translation services, website chatbots, payment integrity, and homelessness predictive services.
“If you take a collaborative spirit, AI can be implemented quickly because you’re cutting out the noise,” Singh said. “There’s a 5X, 10X long-term impact which is always greater than some of the initial pain that I know people feel,” he added.
Singh maintained that increasing AI literacy and education of its impacts is important to its future at HUD, and said increasing knowledge of technology within the department is critical to relationships with the private sector.
“I think standing up the way that people look at technology as a whole has to change. [We need to] reimagine the technology function to integrate AI responsibly and ethically,” Singh said.
“The organizations (local, state, tribal governments, non-profits, and even industry) that receive HUD funding, they’re looking at AI and will eventually leverage AI that could game our funding, or game us,” Singh said. “If we don’t know how it’s being used, and what we can and can’t do, then we are potentially at risk. Those partners are doing things with AI, and we can get smarter. That’s good government,” Singh added.
While Singh left the agency in January, he hopes the new administration will build off the work his team has done, continue to push innovation, and partner with unions in the department’s AI implementation.
“Just having the framework, having people know where to go, what to ask about AI, will enable this new leadership team to invest funding and people behind it,” Singh said. “In the context of current job displacements, vendor contract cancellations, and looming large program cuts from a pending new Federal budget, AI could help, even in the slightest, to buffer some of this impact – to help the millions of Americans that benefit from the HUD mission.”
