Letting AI work in constrained public sector environments

SLMs are built specifically to meet the needs of the department or agency that uses them. Data is stored securely outside the model and is only accessed at query time. Carefully designed prompts ensure that only the most relevant information is retrieved, providing a more accurate response. Use something like Smart search, vector searchand Verifiable source groundcan build artificial intelligence systems that meet the needs of the public sector.

Therefore, the next stage of AI adoption in the public sector may be to bring AI tools to the data rather than sending the data to the cloud. Gartner Forecast By 2027, small professional AI models will be used three times more than LL.M.

Excellent search capabilities

“When people in the public sector hear AI, they probably think of ChatGPT. But we can be more ambitious,” Xiao said. “Artificial intelligence can revolutionize the way governments search and manage vast amounts of data.”

Moving beyond chatbots reveals one of AI’s most immediate opportunities: dramatically improving search. Like many organizations, the public sector holds large amounts of unstructured data, including technical reports, procurement documents, meeting minutes and invoices. However, today’s AI can deliver results from mixed media, such as readable PDFs, scans, images, spreadsheets, and audio recordings, and in multiple languages. All of this can be indexed by SLM-powered systems to provide tailored responses and draft complex texts in any language, while ensuring output complies with legal regulations. “The public sector has a lot of data, but they don’t always know what to do with it. They don’t know what the possibilities are,” Shaw said.

Even more powerful, AI can help government employees interpret the data they access. “Today’s artificial intelligence can give you a whole new perspective on how to leverage this data,” Xiao said. Well-trained SLMs can interpret legal norms, extract insights from public consultations, support data-driven administrative decision-making, and improve public access to services and administrative information. This can help significantly improve the way the public sector operates.

The promise of small language

The focus on SLM shifts the topic from model comprehensiveness to model efficiency. LL.M.s incur significant performance and computational costs and require specialized hardware that many public entities cannot afford. Although some capital expenditure is required, SLMs are less resource intensive than LLMs, so they tend to be cheaper and have less environmental impact.