Google event explores critical intersection between AI and security

Google for Government hosted its Gen AI Live + Labs 2025 event at the tech giant’s expansive event space at Pier 57 in Manhattan Thursday, gathering dozens of senior leaders from public agencies, higher education, healthcare, and more. 

Centering on Google’s cutting-edge innovation in artificial intelligence and how it can be beneficial to the public and the services they use, the conference presented by GovExec and City & State, served as a showcase for what’s next to come in AI and how Google can be a leader in solving real-world issues through it. 

Over the next two years, 94% of state and local agencies are expected to increase AI usage, according to a study from Google and GovExec first mentioned at the event.  Brent Mitchell, vice president of Go-to-Market at Google Public Sector, set the tone of the conference in the opening session, laying out Google’s mission on AI advancement and integration into public life for public good. 

“At Google, we recognize the immense potential of this moment and are deeply invested in ensuring that this innovation benefits all Americans. Our commitment goes way beyond simply developing cutting edge tech, we’re really focused on building a stronger, safer, more competitive America in this AI era,” Mitchell told attendees. He went on to talk mostly about Google’s all-in-one machine learning platform Vertex AI and Gemini, the company’s multimodal generative AI model that can perform, in seconds, all the tasks that would take a human assistant much longer to do. 

Mitchell shared with the audience that the Missouri Department of Social Services, partnering with AI company Quantiphi, leveraged Gemini to streamline the application process for social benefits. They processed 1 million applications, reduced wait times by 70%, totaling 12,000 agent hours saved. 

Similarly in New Hampshire, Google has worked with Rich Lavers, deputy commissioner at the New Hampshire Office of Employment Security, to not only streamline, but personalize the unemployment claims application process to improve customer satisfaction, but also empower workers to take on essential tasks. 

These were just two of the ways Google’s AI initiatives have translated into transforming public agencies and the services they offer. 

What does this look like for agencies relevant to New York, where the event was held?

Google’s Tricia Jamison and Sam Goodgame presented a recent test project in partnership with New York City Transit that utilized Google Pixel smartphones to detect track irregularities with sound waves, all in an effort to catch issues before they become major and make the job of track workers more efficient. 

“NYCT lacked a timely and automated data signal on track nonconformities. In early 2024, they were looking for innovative ways to find them. That’s where our team came in,” Jamison, a Google cloud solutions architect, said. 

She described how her team strapped Pixel phones to the oldest cars in the fleet, using mere duct tape and camping straps. Those phones “collected 335 million sensor readings, one million GPS locations, and 1,200 hours of audio. The data was combined with NYCT’s database of track non-conformities and ingested into a machine learning model running on Google Cloud,” according to the MTA’s press release. 

Goodgame, a senior technical program manager with Google’s Rapid Innovation team, noted how by early detection of issues, it can empower NYCT to accomplish their work more efficiently. 

He presented a real-world scenario that described a track worker in the middle of their shift who spots an issue, but does not have the time to pull out a stocky manual to find out what the issue could be. Instead of trying to hopefully remember the problem for later, in the midst of walking what can be miles of dark tunnel (ultimately delaying the problem’s fix), Goodgame highlighted how something like a manual could be integrated into Google Gemini and workers could speak into a device in plainspeak in real time to receive immediate answers from information only pulled from the manual. This would reduce the time it takes to fix issues by providing immediate, accurate solutions. 

On NYCT’s end, Robert Sarno, an assistant chief track officer and somewhat of a “track whisperer,” helped the Google team get the train moving with their ideas. 

Sarno is an expert at listening to the trains, equipped with a precise ear for hearing nonconformities (Vertex AI detected 92% of track worker-confirmed defects, Sarno’s ear held up a solid 80%). While the Google team of Jamison and Goodgame delivered much of the AI background and application behind the project, GovExec’s City & State spoke directly with Sarno about whether or not this initiative would be system-wide and if its spread could raise fears about track workers’ job security in the future.  Without question, Sarno made it clear that there’d always be “boots on the ground,” stressing the “human element” to the job and regulations about inspection and maintenance. 

On the state level, state Chief Information Officer Dru Rai spoke more broadly with Elizabeth Moon, managing director, customer engineering at Google Public Sector, on the state’s commitment to artificial intelligence and innovation under Gov. Hochul. 

With over a year under his belt, he praised the governor’s championing of AI statewide while touching on some issues the state faces and how AI can help with that, among them tax fraud, building intelligent track management systems, and providing benefits faster, better. 

“We [New York State] ingest 250,000 transactions a minute, so just imagine how would you figure out [the quality of] those transactions without AI?” he asked during a fireside chat with Moon at the event. Rai acknowledged that strictly using AI to deliver public services in place of empowering workers to know how to properly use the technology would result in imperfect service, stressing the importance of upscaling workers to meet the demand.

“We are a service provider. The question is ‘How [do we] train our employees so they can use [AI] and understand the benefits and risks [of using AI]?’” he asked.

The question on just how to train employees who may not traditionally have technical background to be able to properly and ethically use AI in their roles proved a pertinent talking point for much of the conference. Some organizations have the added, and arguably more important, stressor of figuring out how to train the future workforce to be well-equipped to land jobs.

Speaking on a panel about AI in higher education, Ram Ramasubramanian, executive vice chancellor for academic affairs and provost at the State University of New York, noted that “AI has to add value.” Reflecting briefly on a moment during the COVID-19 pandemic where he saw a divide between parents, students, and university administration on the value of technology in (or as) the classroom, he envisioned the future role of AI (and technology at large) as “adding value to education so that there's no question about [the quality of instruction or student experience].”

Jeff Rubin, chief digital officer at Syracuse University and also on the panel, said that he’s used AI in his own lectures to assist students ahead of big exams or projects, a time when demand for his attention far outweighs the supply of time he has in a day to adequately give hundreds of students the attention they may require. This, in addition to numerous other projects that he and university leadership are experimenting with to better utilize spaces on campus, generate student-personalized class recommendations, and beyond.

Chrysoula Malogianni, associate vice president of digital innovation at Old Dominion University, also spoke proudly of her university's longstanding commitment to AI and innovation, touting decades of pushing boundaries with technology, and the debut of AI-driven interfaces outfitted exclusively for the university community. 

Other panels on higher education focused on workforce readiness, entire lines of work being threatened by AI, the gap between curriculums of today and what curriculums should look like tomorrow, and even support for greater collaboration across university disciplines. 

In addition to the panels, attendees were able to take a closer look at some of the projects mentioned throughout the day at side booths, designed to provide hands-on experience using AI.