Five key areas where governments must responsibly deploy generative AI

As detailed in our previous white paper, by 2024, the ongoing digitalization process will further improve the efficiency of government programs and policies. Two important factors driving this digital transformation are data and artificial intelligence (AI). AI plays a pivotal role in unlocking value from data and gaining deeper insights into the wide range of information that governments collect to provide services to their citizens.

With demand for generative AI expected to grow this year, it is essential that the public sector embraces the responsible use of this technology. Only by doing so can government establish itself as a trustworthy steward.

Differences between generative AI and traditional AI

To understand the unique challenges posed by generative AI compared to traditional AI, it helps to understand the fundamental differences. Traditional AI primarily trains models through machine learning, using algorithms and extensive labeled datasets. These models can recognize patterns and adhere to predefined rules to provide recommendations or identify specific actions. For example, existing AI is used to improve the efficiency of spam email filtering, enhance movie or product recommendations for consumers, and enable virtual assistants to help individuals find information.

Generative AI is emerging as a valuable solution for automating and improving routine management and repetitive tasks. This technique excels at applying foundational models, which are large-scale neural networks trained on a wide range of unlabeled data and fine-tuned for a variety of tasks. Effectively identify, summarize, transform, predict, and generate content from large data sets. Implementing this technology in the public sector can significantly improve efficiency, allowing organizations to complete daily tasks with fewer resources.

Generative AI provides unprecedented opportunities to enhance many aspects of government operations and improve services for citizens. This can provide government employees with more powerful tools for answering questions and conducting research. Time-consuming and critical tasks such as creating and managing contracts can greatly benefit from applying generative AI.

Last year, the U.S. Department of State sought feedback on the challenges and security considerations of introducing generative and natural language processing AI into its networks. A State Department request for information in June revealed a goal to improve worker efficiency and accuracy in repetitive tasks related to acquisition planning for market research and contract writing. Generative AI trained with machine learning could help draft new contracts based on this research.

Implementing responsible generative AI

The incredible generative power of this emerging AI technology raises questions about its responsible use in the public sector. For example, contract managers need to know that original research has been faithfully translated into a legally binding contract for two or more parties.

The public has recently encountered generative AI primarily through tools that use existing text, images, video, and audio to create custom content on demand. However, the level of detail regarding training in some of these models may not be sufficient, especially for large corporations or highly regulated industries that rely on public trust.

Developing responsible AI requires government leaders to carefully prepare their internal data to leverage the full potential of AI and generative AI. Setting accountable standards is an important government role, requiring the integration of accountability from the beginning rather than as an afterthought. This includes ensuring the accuracy of AI-generated content and maintaining human oversight to help prevent bias.

Key pillars for responsible AI in government

IBM’s AI development revolves around five guiding principles that help ensure trustworthy AI. Government leaders should prioritize the following factors when considering responsible AI development, training, and deployment.

• equity In an AI system, it refers to the ability to treat individuals or groups fairly depending on the context in which the AI ​​system is used. This means combating bias and preventing discrimination related to protected characteristics, such as gender, race, age, and veteran status.
• seclusion It concerns the ability of AI systems to prioritize and protect consumers’ privacy and data rights while complying with existing regulations regarding data collection, storage, access, and disclosure.
• explainability This is important because AI systems must be able to provide human-interpretable explanations for predictions and insights in a way that doesn’t hide behind technical jargon.
• transparency This means that AI systems must include and share information about how they were designed and developed and what data or data sources were used to feed the system.
• robustness It is the ability of an AI system to effectively handle exceptional situations such as input anomalies. This helps ensure consistent output.

IBM watsonx™, an integrated AI, data and governance platform, embodies these principles by providing a seamless, efficient and accountable approach to AI development across a variety of environments. More specifically, the recent launch of IBM® watsonx.governance™ helps public sector teams automate and address these areas to direct, manage and monitor their organization’s AI activities. This tool facilitates a clear process for organizations to proactively detect and mitigate risks while supporting compliance programs for internal AI policies and industry standards.

As the public sector continues to embrace AI and automation to solve challenges and improve efficiency, it is important to maintain trust and transparency in all AI solutions. Your team must have the ability to effectively understand and manage the AI ​​lifecycle. Proactively adopting responsible AI practices is an opportunity for all of us to improve.

Learn more about how AI can transform government services.