rewrite this title Graph databases are exploding, thanks to the AI boom – here's why

Follow ZDNET: Add us as a preferred source on Google.

ZDNET’s key takeaways

The graph database market, driven by AI, is growing at a rate of almost 25% annually.Graph databases support knowledge graphs, providing visual guidance for AI development.There are multiple dedicated graph database vendors on the market.

Over the past decade, there has been endless churn in technologies shaping the databases behind the applications we run. The rise of NoSQL databases, document databases, and databases built on the web for the web has brought us greater choices.

Also: Gartner says add AI agents ASAP – or else. Oh, and they’re also overhyped

More recently, there has been a boom in the use of artificial intelligence (AI), both in backend systems and through the rise of generative technologies, creating an insatiable demand for databases that can handle and process super-sophisticated workloads. This demand has led to the upsurge of graph databases and knowledge graphs, which are visual databases that can help users manage the requirements of AI. 

Graph databases have been on the rise for several years and now comprise the fastest-growing category within the $137bn annual database market. Again, thank AI — graph databases are seen as the most optimal data backend for AI systems. Spending on these technologies will have a five-year compounded annual growth rate of more than 26%, according to estimates published by tech analyst Gartner at the end of 2024. The overall database management system market will grow 16% annually. In 2025, the Business Research Company projected a compound annual growth rate of 24%.

Also: 95% of business applications of AI have failed. Here’s why

AI demands gobs of both structured and unstructured data, not only fed into applications, but woven into connected patterns that deliver inferences. “The push toward semantic understanding and reasoning in AI systems is something that flat relational databases struggle to support,” said Tony Tong, co-founder & CTO at Intellectia AI. 

Though separate and often confused, graph databases work hand in hand with knowledge graphs — “a graph database is the tool, the engine for identifying connections within a given dataset. A knowledge graph is the representation of the data itself — the product of a graph database,” said Daniel Bukowski, chief technology officer at Data². 

Also: 71% of Americans fear that AI will put ‘too many people out of work permanently’

“Knowledge graphs provide AI systems with real-world information and how that information is related, which helps the AI answer questions with more accuracy and nuance. Graph databases allow you to search through data more efficiently and provide context not found in raw data alone.”

Graph environments can be applied to functions, such as real-time analytics, fraud analytics, retail, and logistics, said Shalvi Singh, founder of Healthengine.us, and senior product manager at Amazon AI: “Knowledge graphs are aiding large language models by offering ample context for structured reasoning and by enabling contextual understanding.” 

The ranking of the most popular graph databases includes the following technologies (source: DB-Engines):

Neo4j Graph Microsoft Azure Cosmos DB Aerospike ArangoDBOrientDB GraphDB Virtuoso Amazon Neptune Memgraph GraphNebulaGraph

Of course, implementing graph databases is not an overnight project. For example, “incorporating data from different sources is still subject to inconsistency or out-of-date information,” Singh cautioned. 

Also: AI agents will be ambient, but not autonomous – what that means for us

Scalability is also an issue, as the performance of these data environments may deteriorate as datasets increase in size and complexity. “These technologies do not replace traditional databases,” she added. More hybrid arrangements may be necessary for scalability purposes.

Plus, graph databases and knowledge graphs “often require specialized expertise, detailed planning, and careful structuring of interconnected data,” said Bukowski. “Even though knowledge graphs have been used for decades, graph databases are a newer, fast-growing segment of the database market, meaning that it can be difficult to obtain, implement, and master both of these tools.”

Also: Gen AI disillusionment looms, according to Gartner’s 2025 Hype Cycle report

Without data, there can be no AI. For those looking to provide greater data support for their AI efforts, graph databases and adjoining knowledge graphs represent visual connections that assure more on-target AI efforts.