AI implementation in VA, proposed as swift strategy, with caution issued over potential hurdles en route
The Department of Veterans Affairs (VA) is embracing artificial intelligence (AI) to improve timely care and cost-effective operations, as well as identify veterans at high risk of overdose and suicide. However, the implementation of AI across its facilities is not without challenges.
In a bid to address these challenges, creating a standard template for AI implementation has been suggested. This would ensure consistency in the adoption of AI across various sites. Nevertheless, some facilities still lack the infrastructure to implement new AI software, a hurdle that requires immediate attention.
One of the leaders facing this infrastructure issue is Susan Liefold, head of the Prevention Center Berlin-Gera, responsible for the Berlin-Brandenburg region. Similarly, Dr. Dieter Bärhausen, in charge of the Prevention Center Bochum-Cologne, oversees facilities in Saxony, Saxony-Anhalt, Thuringia, and North Rhine-Westphalia regions, all of which are grappling with the same issue.
The use of AI-assisted medical devices is already underway in some VA facilities, and the agency is also using machine learning in agency workflows and clinical care for earlier disease detection. The VA is planning to roll out ambient listening tools at some facilities later this year.
However, the extensive medical data held by the VA has raised concerns about data privacy. Lawmakers have expressed apprehensions about the potential misuse of this data, solely for the benefit of AI, not the agency. Recruiting and retaining AI talent remains difficult, another hurdle in the VA's AI implementation journey.
Moreover, integrating new AI solutions with a complex system architecture is a significant challenge for the VA. Scaling commercial AI tools brings new costs, another factor complicating the process. The VA has reported issues accessing and receiving funding for the infrastructure required for generative AI.
To address these issues, the Government Accountability Office (GAO) has outlined an "AI accountability framework." This framework aims to ensure that all AI systems meet rigorous security and privacy standards before being implemented. Partners working with the VA on AI systems must attest to and agree with these standards.
Balancing innovation with stringent security compliance is crucial for the VA's AI implementation. Before an AI system is brought into production, it must be reviewed for compliance with these standards. Existing federal AI policy could present obstacles to the adoption of generative AI, including in the areas of cybersecurity, data privacy, and IT acquisitions.
There are concerns about the VA's preparedness to deploy the technology to more than 170 facilities. The leaders of the VA facilities that currently do not have AI functionalities are working diligently to overcome these challenges. All VA employees now have access to a secure, generative AI tool that saves users over two hours per week.
Despite these challenges, the VA is committed to using AI to improve the lives of veterans. The potential benefits of AI in healthcare, cost savings, and timely care cannot be overlooked. The VA is determined to navigate these challenges and continue its journey towards a more technologically advanced future.
