Federal agencies face a range of tasks as they step up their use of artificial intelligence, including the need for “hiring or training employees who understand and use the technology responsibly,” says a report from the Partnership for Public Service.
“The federal government should emphasize expertise in technical, digital and data skills. It should provide extensive and ongoing training to employees so they can create, understand, manage and work with AI technology,” says the report, done with the IBM Center for the Business of Government.
The report notes that federal training can suffer from shortfalls caused by budgetary and other reasons but says that “at the outset, even small changes, such as educating employees on key AI terms and definitions, could be beneficial.”
Oher management-related challenges in incorporating AI include:
* Quality of Data—”Knowing that biased data may lead to biased results, agencies need to pay special attention to what information is being used with these new technologies. AI technology is “trained” on data, yet not all the information that has been collected over the years is necessarily of the highest quality.”
* Security—”AI is vulnerable in several ways if designed without proper security measures. Attacks could alter AI training data or introduce corrupted or incorrect data that changes the conclusions of the AI tool. Hackers also could act to reveal personally identifiable information in the data on which an AI tool was trained.”
* Transparency—”Lack of transparency can pose issues when people want an explanation for why decisions were made . . . Without clarity about how AI produces its recommendations and conclusions or understanding from employees as to how to explain results derived from AI technology, governments may risk losing the public’s trust.”
* Acquisition—”Outdated federal acquisition and budget processes prevent agencies from buying and deploying new technology quickly and efficiently . . . The rapid pace of AI development and improvement can leave government lagging behind, as has been the case with the introduction of many emerging technologies in the past.”