To anyone following the AI space, it should come as little surprise that public sector agencies at the international, federal, state, and local levels are making significant investments into AI technologies across all the patterns of AI applications. From AI enabled chatbots, to predictive analytics, to autonomous systems of all sorts, to process automation, and many other use cases among the seven patterns of AI, public sector agencies and organizations are pushing forward with AI to improve service to their stakeholders, reduce the cost and time of providing services, and provide greater consistency in operations.
Some agencies in particular have been making great strides when it comes to AI adoption. The US Department of Health & Human Services (HHS) has been very forward thinking in their adoption of transformative technologies including AI and blockchain. In this article, Oki Mek, Senior Advisor to HHS CIO at HHS shares how HHS is adopting and implementing AI, how adoption of new technologies in the public sectors differs from the private sector, what the HHS Reimagine project is all about, and what steps the federal government can take to attract and retain skilled AI talent.
How is HHS currently adopting AI technologies?
Oki Mek: Artificial Intelligence is such a broad term and at the current stage I refer to AI as “Augmented Intelligence”. In the federal space, I see AI serving in an assistive role to enhance human intelligence and decision making. At HHS, we are primarily focusing on supervised and unsupervised Machine Learning, to gain better insights into big data.
How are these ways unique or similar to other agencies’ adoption of AI?
Oki Mek: I see other federal agencies beginning to adopt Machine Learning, as the federal space is currently experiencing a big data paradigm shift. There are numerous benefits in using AI to glean insights into data; descriptive analysis allows for better presentation of the data and can identify gaps and anomalies. Furthermore, it provides transparency and visibility that provides better overall service delivery. This is important for agencies because it closely aligns with the President’s Management Agenda (MGT), Federal Information Technology Acquisition Reform Act (FITARA), and Modernizing Government Technology (MGT) Act. Better data insights also improves risk management when dealing with audits around data discrepancies such as the Data Act and GAO Audit. Machine Learning also provides a great tool to conduct predictive analysis to understand risk and plan accordingly, and in some cases provides informed suggestions. Simply put, you can’t manage something you can’t measure.
What are some of the unique challenges HHS has with regards to data and data collection?
Oki Mek: User agreement between or among the parties is a challenge. Everyone wants to gain better insights into their data but knowing where to start can be daunting – from administrative policies such as privacy and security, to business considerations and value, to liability concerns.
What are some ways in which the government is learning from industry and the private sector on AI adoption?
Oki Mek: The best approach to modernization in the federal space is partnership with small and large businesses. Collaboration is the key, not just with industry but with other federal agencies. There are a great deal of smart people in both public and private industry who bring lessons learned and best practices that could be leveraged in agencies’ efforts to modernize. We have had a lot of success with our small business partners because they are willing to take more risk and deliver more with less.
What are some of the challenges you’ve seen with AI adoption in your agency and how are you overcoming them?
Oki Mek: Federal agencies want AI but the challenge is the data. You must develop a strong business plan that covers a 360 degree approach. Also, managing risks and expectations of AI implementation are challenges. Trust and security will play a major role in AI algorithms; that’s where I believe Blockchain can alleviate the trust concerns by providing transparency, immutability and visibility into the algorithms that will provide recommendations to the decision makers. Trust in the algorithms is paramount considering that the recommendations may affect whether someone gets a job, wins a contract, or is awarded a grant. The public must have full transparency into the criteria within the algorithms to avoid unexpected bias or unauthorized changes. This closely aligns with the White House Executive Order 13589, to establish three goals for AI: ensure public engagement, limit regulatory overreach, and promote trustworthy technology.
What are some interesting or surprising insights you can share about HHS’s use of AI?
Oki Mek: We learned that when there is sufficient data, the AI predictions can be fairly accurate. For example, AI can parse through millions of records and identify fraud, waste and abuse with a relatively high degree of accuracy. HHS is currently using AI to conduct pre-award risk assessments of COVID grants.
What is the HHS Reimagine project and how does this tie into AI efforts?
Oki Mek: The ReImagine HHS initiative helped bring emerging technologies like Blockchain, AI, and RPA to HHS. Several notable innovations were started from ReImagine HHS in acquisition, grants, and mission. ReImagine HHS identified six strategic shifts as opportunities for the department to advance its work: Leveraging the Power of Data, Restoring Market Forces, Putting People at the Center of HHS Programs, Generating Efficiencies through Streamlined Processes, Making HHS a More Innovative and Responsive Organization, and Moving to a 21st Century Workforce.
Looking more broadly, what are a few areas where you’re seeing AI effectively being used in the federal government?
Oki Mek: AI can be broadly applied in the federal government in areas like cybersecurity, fraud, risk, compliance, audits, etc. For example, AI can be applied to detect fraud in payments, identify security vulnerabilities, monitor compliance, etc.
What can federal agencies do to attract the skilled workforce it needs to keep up with technological innovations?
Oki Mek: Federal agencies are unlikely to be able to hire and retain AI talent because there is an immense shortage in supply. Consequently, the best strategy is to partner with industry.
What AI technologies are you most looking forward to in the coming years?
Oki Mek: Cybersecurity AI along with Blockchain. The Authority to Operate and FedRamp processes take about 8-12 months. By leveraging emerging technologies such as AI and Blockchain, these processes can be reduced to 3-4 weeks. Furthermore, security compliance, change/configuration management, and DHS Continuous Diagnostics and Mitigation (CDM) have to merge to achieve the true essence of real-time risk management.