Friday, March 6, 2026

Va Chief Ai Officer: Technology is secondary to solve human problems

Share

Last year, Charles Worthington, then technology director at the US Veterans Department, was responsible for supervising all artificial intelligence efforts throughout the department.

In addition to the CTO Worthington title, he got a modern one: Chief Officer AI.

Worthington’s background is software creation and product design. He sums up this as “using computers for solving problems”. And so he and his team dealt with work at VA.

“This ability to attempt to determine how to best solve technology important business problems or problems that VA has is probably the most important skill,” said the leader of Va IT. “The technology itself is a secondary problem. It is more essential to determine the most essential problems to solve.

“When you find out what these problems are, you can be creative to find out what is the best way to use technology to solve them, regardless of whether this technology is artificial intelligence, or more traditional software or anything else,” added Worthington.

Government digitization

Worthington helps the government to accept the best practices of state-of-the-art digital services since 2013. He is currently responsible for coordinating the exploit of AI by VA, while promoting innovation and risk management.

He also manages the greater technological strategy of the Agency and conducts a team of technical experts who exploit the best practices of state-of-the-art technologies to improve VA services. He conducts efforts to improve the experiences that veterans have during interactions with VA online, improving the processing of benefits, increasing the interoperability of health data, modernization of VA data analysis possibilities and assistance in the IT Technology Office in the adoption of the best practices of state-of-the-art engineering.

Before joining VA in May 2017, Worthington helped create digital services in the USA. He is a co -author of the “PlayBook of Digital Services” and conducted recruitment efforts for over 150 software engineers, designers and managers from the best technology companies. He helped in the implementation of the involvement of digital services in the services of citizenship and immigration of the United States, a diminutive company administration as well as transport and interior departments.

But today Worthington is in VA. And the roles of Cto and Caio keep him quite busy.

“Part of my work as a CTO is to ensure that VA identifies and uses new and developing technologies to help in the implementation of its mission,” Worthington explained. “Therefore, the scope of my role was natural for the main duties of an AI officer, which consists in helping to accelerate the reception of VA This specific, high impact on some of our most essential problems.

“The synergy between the existing CTO mandate here in Va and the duties of the director of AI in the government was a good fit, which made it a natural decision to appoint CTO to act as the director of AI in the agency,” he added.

Working with CIO

Worthington reports to his office. The IT office deals with technological needs for the entire department, which of course covers the great administration of healthcare, but also brings the benefits of administration, administration of cemeteries and staff offices, including human resources and finance.

“We really focus on this, starting with the problems that our clinicians have, our veterans, our patients and how to use the technology to help these problems,” he said.

“There is one large string that basically accelerates AI’s reception in specific, large areas, will most likely help Va help solve some of its most important problems,” he contradicted. “I spend a lot of time trying to identify these problems, and then find out what, if at all, stops us from using artificial intelligence to solve them.”

On the one hand, his role is to be a master of AI and to make sure that VA creatively thinks about the possibilities of successful exploit. On the other hand, the role is responsible for the management of AI.

“In this way, VA ensures that the artificial intelligence we use is safe, effective and trustworthy, and basically does what we think and we are sure that our ability to monitor and measure these things,” explained Worthington. “And at VA, the director of AI is also responsible for ensuring an effective management process of the agency.

“So these are two parts of the work on which I spend a lot of time,” he added.

Helping colleagues thinking about artificial intelligence

A typical day for Worthington may include a meeting with employees somewhere in VA who are working on an essential problem and discussing how AI can be used to assist – seeing ideas that they are already working on and what they may need or what they may need to unlock something.

Or maybe put in their team members or some of his team ideas so that a group of Worthington people visits to think about solving a problem with AI, which they may have not yet considered.

“And then On the side of governance, this is really a lot of communicating with interested parties, people who run different parts of VA, who have artificial intelligence in their areas, making sure they understand their requirements, and that they have a good plan to measure this effectiveness and credibility so that we can be sure that our use cases meet our high standards, “he noted.

When it comes to healthcare, VA currently has over 100 cases of exploit in production, which covers everything, from imaging to radiology to pathology. Many of these cases of exploit are often associated with medical devices approved by the FDA, which have imaging elements in them.

“We also have what you consider to be more traditional machine learning models that make forecasts,” said Worthington. “For example, they generate a factor or forecast of risk assessment and using it to identify specific patients who can be good candidates for specific types of care. We have many examples in the field of forecasts or risk assessment.

“And then we have many cases of use focusing on general performance in business operations,” he continued. “For example, we are dealing with a pilot of a script in the environment. And we exploit generative artificial general intelligence, chatgpt experience, which allows different types of personnel to exploit a immense language model with VA data to assist them answer E -Maile or view the rules of principles or browse a immense amount of information.

Lots of cases of exploit of healthcare

There is a huge amount of artificial intelligence in VA in VA in VA, and cases of exploit lead a range.

“Decision support, imaging, Natural language processing is quite common categories, “said Worthington. “A VA publishes our inventory of cases of exploit of AI in principle every year, every year, Recently in December. We are talking in detail about all these cases in public. Everyone can go to our website and see the full supply of AI exploit.

“I am proud of this because I position VA as one of the most transparent health care systems in the country in terms of the use of artificial intelligence,” he continued. “I know people in many healthcare systems and I am sure that many people you talk to are struggling with the inventory of their use by AI and how to rule it. So I am proud that VA publicly shares our comprehensive list of all different cases of AI, which we have in production.”

In the space of medical devices, VA has published some studies on the effectiveness of the AI ​​imaging tool, which she uses to support colonoscopy. It is called GI Genius. VA implemented it in over 100 facilities and published research that shows that the AI ​​tool reduced the number of polyps omitted in colonoscopes by up to 50%.

“So I am excited not only because we implemented it, but also because of the research,” Worthington delighted.

And helping in suicide veterans

Another example is the risk assessment category/decision support.

“We have developed an algorithm called Reachvet, which uses artificial intelligence to identify veterans who are at the highest risk of suicide,” he said. “We are able to exploit this risk assessment algorithm to direct additional interventions to the most endangered patients.

“We published research that shows that the model was associated with a decrease in mortality for any reason,” he concluded. “This is not only the model itself, but the interventions that we aim based on the results of this model. So of course AI is not the answer itself. It is when you combine analysis and risk assessment with the entire healthcare program that is able to act in these insights.”

Another example of Worthington is excited by the early internal version, Generative tool and a general purpose. It is called VA GPT, available for exploit in the VA network. And thanks to the VA security position and VA business agreements with suppliers, the Agency can allow users to exploit VA data.

GPT for everyone in VA

“This is basically a chat interface that allows clinicians or any VA employee to interact with a large language model to help them at VA work,” he explained. “We currently have over 40,000 VA GPT users. And a survey of these users is really stimulating, over 80% of them agree that the tool has increased their efficiency and report on an average of three hours a week to the tool user.

“This shows the potential of some of these technologies to generate many time savings,” he added. “I am more glad that people report increased satisfaction from work when they have access to these tools.”

A really powerful sign that AI is on the right track successfully supports clinicists and VA employees to perform things to which they are trained, act at the top of their license and exploit computers to assist in some things that otherwise could be a hardworking work in the field of healthcare, he said.

“It can be a really powerful multiplier of strength for our employees,” Self -Smotton summed up.

Watch now: How to launch the AI ​​Healthcare project, according to the head of VA AI

Latest Posts

More News