5 practical concepts for building trust in government digital strategies with Elastic

In a previous blog post, we explored some recurring concepts in the digital strategies of European countries. Here, we’ll dig a little deeper and explore how Elastic can help the public sector meet these priorities with efficiency and security. 

We’ll explore five strategic concepts that directly align with the pillars in the Netherlands' Digital Strategy (NDS)¹ but are also reflective of other countries’ strategies:

• Open source and open standards

• Sovereign infrastructure

• Responsible AI

• Cybersecurity and resilience

• Data driven government

Trusted globally and headquartered in the Netherlands, Elastic helps Dutch and European public organizations — ranging from ministries and regulatory bodies to municipalities and executive agencies — gain visibility, better control, and insight over their data.

Public organizations have a big responsibility when working with artificial intelligence and machine learning (ML). They need to provide transparency around its use while protecting sensitive data. This is not new, but the rise of generative AI (GenAI) made it much more apparent. Fortunately, Elastic has been active in the AI field for close to a decade now. And our platform has been built with data protection in mind and has evolved to integrate AI responsibly and securely. For example, all AI training performed in Elastic has automatic governance applied; using the model map attached to each trained model, we give complete insight into what data was used, how the model is used, and who has used it.

When public organizations use AI, they are expected to have full explainability about the results from the AI models. For regular ML models like regression or classification, Elastic offers the possibility to train those models with “explainability” in mind. This way, any inference from those models comes with concrete justifications from the data. This simple, basic capability of Elastic can make AI usage safer.

Neural networks that are the basis of all natural language/semantic search and the large language models (LLMs) that underlie GenAI are unfortunately nondeterministic by nature; it is not possible to “explain” this class of AI models. There is simply no practical way to tell why they take the logical steps they take. But some models are more open to showing their internal “thinking” strategies.

Elastic enables users to choose any model they want based on criteria they find important. We have a full suite of connectors that make it easy to connect to any publicly available model service. And, especially important for public organizations, we can connect to locally hosted LLMs. As more models become available for download, a major privacy hurdle keeping many public organizations from exploring the advantages of GenAI and retrieval augmented generation (RAG) architecture is finally falling away.

The combination of AI and data protection by design and the capability to run an end-to-end AI pipeline completely in your own trusted environment can simplify the process of operationalizing data and AI in public organizations.

Now, of course, when you start ingesting more data to feed the security use cases, your Elastic environment can be adjusted in a cost-effective way to support the growing volume and maintain mission-critical operations. For example, instead of adding more nodes, you can choose to retain specific data sets for a smaller amount of time. For security (and especially compliance) use cases, a retention time of 10 to 15 years or more is often a requirement. But datasets that are only used for monitoring like system metrics can often be deleted much sooner, freeing up space in your existing environment.

Long-term data retention is a standard best practice for security use cases. And with Elastic’s flexible tiering approach, keeping decades of data in low-cost blob-storage is possible without going over your budget. Even in Elastic’s frozen tier, data is still fully searchable and ready for analysis without the need for a time-consuming ingesting, thawing, or importing.

Continue exploring the topic

• Blog: Building the foundation of trust in government digital strategies 
• Blog: Understanding data mesh in public sector: Pillars, architecture, and examples
• Blog: Understanding AI in government: Applications, use cases, and implementation
• White paper: Cybersecurity Guide for Public Sector: Protecting data and assets in the era of AI & efficiency⁣
• Report: Analyzing the economic benefits of Elastic Security

Sources

1. The Netherlands' Digital Strategy, Netherlands Digital Government, June 2025 

2. Elastic Netherlands User Group, Meetup

The release and timing of any features or functionality described in this post remain at Elastic's sole discretion. Any features or functionality not currently available may not be delivered on time or at all.

In this blog post, we may have used or referred to third party generative AI tools, which are owned and operated by their respective owners. Elastic does not have any control over the third party tools and we have no responsibility or liability for their content, operation or use, nor for any loss or damage that may arise from your use of such tools. Please exercise caution when using AI tools with personal, sensitive or confidential information. Any data you submit may be used for AI training or other purposes. There is no guarantee that information you provide will be kept secure or confidential. You should familiarize yourself with the privacy practices and terms of use of any generative AI tools prior to use. 

Elastic, Elasticsearch, and associated marks are trademarks, logos or registered trademarks of Elasticsearch N.V. in the United States and other countries. All other company and product names are trademarks, logos or registered trademarks of their respective owners.