The professional world largely accepts the idea that generative artificial intelligence (GenAI) has the potential to transform and improve certain kinds of work. However, numerous government agencies and employees are reluctant to embrace the GenAI push. In general, government employees tend to be more skeptical than their private-sector counterparts about the impact GenAI could have on their work. They’re more careful about expressing optimism about AI’s role in government over the next few years.
In the recent Thomson Reuters Generative AI in Professional Services Report, most government officials in legal, law enforcement, and risk management roles believe GenAI can be applied to industry work. Government workers also said that using AI more in their departments would probably be hard because of problems with administration, budgeting, and culture. Some government departments even have policies that prohibit the use of AI.
Benefits vs. risks of GenAI use in government
This antipathy toward AI is understandable, but also a bit ironic, because high on the list of institutions that could benefit most from the emerging capabilities of GenAI are federal, state, and local governments.
People often feel disenchanted with the government because it’s too slow, unresponsive, or bogged down in red tape. GenAI has the potential to change these dynamics. Techno-optimists think GenAI can make government services better and more responsive. They also want it to be more efficient.
We must recognize that the private and public sectors are quite different in terms of how they view data security, assess risk, and allocate budgets. Also, governments are often too slow to adopt new technologies for good reasons. When it comes to using GenAI in government, officials should weigh its potential benefits against the risks and limitations of sticking to traditional methods.
The GenAI difference: improving the capabilities of machine-learning
GenAI is a type of artificial intelligence that greatly improves the capabilities of traditional machine-learning AI. For example, machine-learning AI can automate repetitive tasks, searching large databases, and organizing information. GenAI, on the other hand, uses much larger datasets and technology that works more like the human brain. This development allows GenAI to create new ideas, concepts, and insights, including images and videos. GenAI can also understand and respond to spoken questions or commands, and learn to fix problems, correct errors, and improve its results.
When embedded in the same system, such as the government, machine learning and GenAI can complement each other. This combination creates a form of computing that’s far more useful, efficient, flexible, and responsive than previous technology. In fact, GenAI is so useful for governments that it could lead to a complete re-thinking of how government services are managed and delivered.
GenAI’s conversational abilities make it easier for government professionals of all skill levels to use AI systems. This means they can work with AI systems without needing the coding skills that product engineering types usually need. Such user-friendliness makes GenAI a valuable tool for improving government services and operations across different roles and departments.
Government agencies adopt chatbots to overcome manual shortcomings
One of the causes of inefficiency in government systems is the volume of data agencies must gather and analyze. Manually inputting data significantly slows down the process. Inadequate search technologies make it even harder to find and use relevant information, further delaying things.
One result of this cumbersome process is that government employees spend too much time searching for information they don’t have readily available. Similarly, when citizens call agencies like the Internal Revenue Service (IRS), Social Security Administration, or Department of Health and Human Services, they face longer wait times for answers.
The current system is flawed. The problems are so deeply rooted that the public accepts them, along with the frustrations, believing this is just how the government works. But it doesn’t have to be this way.
In December of 2019, many federal and state government agencies began using voice-enabled chatbots to answer basic questions about government services. Most of these bots use machine learning and natural-language processing (NLP) to understand text or voice queries. They are programmed with answers to a few hundred of the most frequently asked questions. If the bot can’t answer a question, it usually refers the matter to a human support person.
These machine-learning and NLP-based chatbots are typically limited by their programming. They’re unable to answer questions that haven’t been pre-loaded. When combined with GenAI, chatbots have a much more extensive range of capabilities.
How GenAI chatbots can improve government functions
GenAI-enabled chatbots can understand and respond to almost any language in a more conversational way. They can understand what someone’s trying to ask, even if the question itself is confusing. These chatbots can remember previous conversations, ask clarifying questions, and predict answers based on the context of the conversation. They can make personalized recommendations, and assist people with filling out forms, scheduling appointments, paying bills, and more.
Behind the scenes, GenAI chatbots can analyze citizen feedback from surveys and social media. They can identify areas for improvement and automatically refine their answers over time as they learn more about the questions they receive. In short, GenAI chatbots can provide a much more personalized, natural-feeling interaction, and provide people with a much larger universe of answers and guidance.
The effectiveness of a GenAI chatbot, or any chatbot, depends on the quality of its data and programming. GenAI can improve many applications to get great results. This could change how government services are given out in many areas.
With that in mind, consider how GenAI might improve some of these common government functions.
Public safety
Problem. Emergency responders need to assess the nature and severity of incidents quickly, but they only know what the dispatcher tells them before arriving at the scene.
GenAI solution. While responders are on their way, GenAI analyzes data from various sources such as traffic and surveillance cameras, social media, police scanners, and body-cam footage. It provides real-time updates of disaster scenes, including photos and video footage, enabling faster and more strategic responses during a crisis. GenAI also examines crime data patterns and predicts threats to help governments develop emergency plans.
Urban planning
Problem. Urban planners prioritize construction projects and development plans years in advance, but they often lack the information needed to optimize their resources and budget.
GenAI solution: GenAI can synthesize citizen feedback, analyze past resource deployment, assess past and projected weather and traffic patterns, and create visual models of plans. It uses demographic data and housing scenarios to propose the best approaches for the government’s time, budget, and resources.
Human services
Problem. Human service caseworkers are overwhelmed with applications for various services, but they can only do so much in a day, which delays the delivery of those services.
GenAI solution. GenAI can help caseworkers by sorting applications based on need. As cases move through the system, GenAI pulls data from applications and generates personalized status updates in the recipients’ preferred language, speeding up communications and the application process.
Tax collection
Problem. Identifying potential cases of tax evasion and fraud requires a lot of time, expertise, and money because of the vast amount of data involved.
GenAI solution. GenAI can analyze vast quantities of data and identify patterns, inconsistencies, and trends from current and past tax returns as well as other publicly available documentation. It then flags suspicious activity and recommends next steps based on pre-established protocols.
GenAI improves government efficiency using insights from large datasets
GenAI scenarios need technology and communication systems that government workers don’t have yet, but they can get them soon. GenAI is already improving government operations by improving existing apps and using data the government already has.
Many government agencies are using a combination of machine-learning AI, GenAI, and human oversight to:
- Automate and streamline workflows, review documents, and extract hard-to-find data
- Summarize and analyze large reports
- Screen job applicants and automate the onboarding process
- Automate parts of the contract bidding and procurement process
- Help citizens navigate bureaucracy through chatbots and virtual assistants
- Assist with compliance and reporting obligations
- Identify fraud, waste, and abuse
- Fight cybercrime and track criminal money trails
- Draft basic legal documents, contracts, and policy statements
Government workers aiming to accomplish more in less time can use GenAI to:
- Speed up the bidding and contract process
- Accelerate and improve all kinds of scientific research
- Provide instant policy summaries to lawmakers
- Inform policymakers with data-driven insights and analysis
- Forecast risks, trends, and threats of all kinds, including economic and military
- Predict future needs for infrastructure, housing, food, health services, etc.
- Optimize budget and resource allocation
- Improve compliance and overall government productivity
In these cases, governments benefit from AI’s ability to quickly process and analyze vast amounts of data. GenAI enhances this capability by generating unique, actionable insights from data in a more efficient and user-friendly manner.
Thomson Reuters enhances CLEAR with new GenAI capabilities
Thomson Reuters has been a pioneering leader in the AI field and is committed to remaining at the forefront of GenAI technological research and development. One of their products, CLEAR, an industry-leading investigative software, now incorporates new GenAI capabilities. Many government agencies use CLEAR for vendor due diligence, fraud prevention, and law enforcement.
CLEAR is a powerful tool that generates an enormous amount of data. The software’s dashboard, search tools, and alert systems provide a great deal of insight. However, vital information can sometimes be overlooked. This isn’t due to the software’s depth of data and analytical capabilities. Rather, users might not know how or where to find the information.
The new GenAI-powered Risk Analysis Summary (RAS) in CLEAR
To help government users get the most out of their investigations, CLEAR now includes a GenAI-powered tool known as RAS. This enhancement is designed to reduce the time it takes CLEAR users to conduct a search and evaluate results.
When a government agency conducts due diligence on a potential new vendor or supplier, RAS now includes a GenAI-powered risk overview and summary at the top of every report. The summary uses both open-source and proprietary data to clearly highlight the report subject’s overall potential risks.
In general, an RAS report summary typically includes a:
- Company overview. A narrative description of the business’s operations.
- Risk overview. A narrative description of potential risks.
- Key business details. A list of legal names, addresses, filings, ticker symbols, and information on essential personnel and beneficial owners.
- Adverse records and media review. A list of any adverse records, media events, or sanctions mentioning or involving the business in question.
- Social media activity. Links to assess a company’s social media presence.
RAS gives a short overview of all important information about a company. It also warns users about any risks or issues that might need more research. RAS also includes links to relevant source documents for further clarification, if needed. Additionally, it offers source attribution for all statements made in the AI-driven portions of the report.
RAS is just one example of how GenAI can enhance the capabilities of an existing tool or application. More importantly, it shows how GenAI technology can help government officials extract useful insights from large amounts of unruly data, saving time and money while delivering more effective results.
Develop ethical GenAI guidelines to boost data privacy and security
GenAI technology can be used by many government agencies, but there are many reasons why governments don’t use new technology quickly. Policy roadblocks, budget limitations, bureaucratic inertia, and the sheer size of many government agencies all play a part. Additionally, a lack of technically skilled workers and aging technological infrastructure contribute to the issue. However, two of the chief concerns regarding the use of AI in government are data privacy and security.
Government systems store vast amounts of personal data, making data security essential. GenAI technology requires access to a broad range of data sources to work properly, including databases full of private, highly sensitive information in a government setting.
Without robust data-protection measures, unauthorized access is a high risk, especially since government systems that house sensitive personal data are prime targets for hackers and fraudsters.
As government officials weigh the benefits and risks of AI, they must establish transparency and accountability systems and strengthen security protocols. They also need to develop strategies and guidelines for the ethical use of AI and integrate these into employee training.
Embrace the power of GenAI with CLEAR for effective governance
GenAI is here to stay in government agencies and will only improve over time. As technology advances, the private sector is rapidly adopting GenAI, transforming work and business with compelling benefits.
Thomson Reuters is committed to the safe and secure deployment of AI technology in both government and business. They integrate transparency about AI use in their products from the start. The future of commerce and governance depends on this commitment. Without awareness of the risks and a comprehensive strategy to address them, the potential benefits of incorporating GenAI into government processes could be undermined. Unintended consequences can arise when policymakers don’t fully understand the technology.
GenAI will become more common in government. It’s important for society to keep up with its growing technological skills as it tries to govern.
When used wisely and with proper guardrails, GenAI can make governance more effective, efficient, and responsive. Pairing it with machine learning and human expertise boosts these advantages, using the strengths of both humans and technology. GenAI might even help restore some of the trust in government institutions that has declined in recent years.
Discover how CLEAR’s GenAI features can contribute to more effective and trustworthy governance.