Anyone who has experimented with artificial intelligence in an educational setting has likely encountered the "context problem." A foundation model might be highly capable of generating text or writing code, but when asked a specific question about an institution's curriculum, a departmental policy, or the nuances of a specific vocational training module, it often falls flat. The AI lacks the internal context that human educators and administrators rely on every day. To address this, organisations have spent the last year building bespoke "knowledge bases" to feed their AI agents, resulting in a fragmented landscape of incompatible formats, proprietary systems, and siloed information.

Recently, Google Cloud introduced a potential solution: the Open Knowledge Format (OKF). This open specification is designed to standardise how organisational knowledge is packaged, shared, and consumed by both humans and AI systems. Rather than relying on complex new databases or proprietary software development kits, OKF takes a refreshingly simple approach. It formalises the "LLM-wiki" pattern into a portable format using familiar, accessible technologies: standard Markdown files enriched with YAML metadata.

My colleague George Bekiaridis recently highlighted this development on LinkedIn, pointing out why this shift is so crucial. As he noted:

"Google Cloud recently introduced the Open Knowledge Format (OKF), an open standard designed to improve how organisations, teams, and artificial intelligence systems share and utilise knowledge. Rather than exchanging only files or datasets, OKF enables the sharing of the essential context behind them: metadata, definitions, relationships between concepts, documentation, and operational knowledge."

The brilliance of OKF lies in its simplicity. An OKF bundle is essentially a directory of Markdown files. Each file represents a concept—whether that is a dataset, a rubric, a policy document, or a technical manual. The YAML frontmatter at the top of each file provides structured, queryable fields like title, description, and tags, while the Markdown body contains the actual content. Because these files can link to one another, they create a rich, cross-referenced graph of knowledge.

As George further explained:

"What makes this initiative particularly interesting is its simplicity. The format is based on familiar technologies such as Markdown files enriched with YAML metadata. It does not require specialised platforms or complex infrastructures, making it easy to create, maintain, share, and integrate across different tools and environments. Its goal is to establish a common 'language' that can be understood by both humans and AI-powered assistants and agents."

For the vocational education and training (VET) sector, this development is highly significant. Educational institutions generate vast amounts of contextual knowledge: assessment criteria, pedagogical guidelines, industry competency frameworks, and administrative workflows. Currently, if a VET provider wants to build an AI tutor or an administrative assistant, they have to manually extract this knowledge from various learning management systems, shared drives, and policy manuals, then reformat it for a specific AI vendor's platform. If they change vendors, that work is lost.

OKF offers a way out of this vendor lock-in. By adopting an open, vendor-neutral standard, educational institutions can curate their institutional knowledge once. This "wiki" of knowledge can live in standard version control systems, be updated by human staff, and be read natively by any AI agent that supports the OKF standard. It separates the creation of knowledge from the consumption of it.

The challenge facing education today is rarely a lack of data. The challenge is ensuring that data is understandable, contextualised, and interoperable. George captured this perfectly in his concluding thought:

"For professionals working in education, research, data analytics, or artificial intelligence, this development is especially significant. One of the biggest challenges today is not the lack of data, but the lack of shared standards for describing, contextualising, and understanding it. The Open Knowledge Format represents an important step towards a more interoperable, open, and collaborative knowledge ecosystem, where information can be effectively leveraged by both people and machines."

As AI continues to integrate into educational workflows, the tools we use to manage our institutional memory must evolve. Initiatives like the Open Knowledge Format demonstrate that the future of AI in education does not necessarily require adopting complex, opaque systems. Sometimes, the most powerful solutions are those built on open, simple, and shared standards.

References

[1] McVeety, S., & Hormati, A. (2026, June 12). Introducing the Open Knowledge Format. Google Cloud Blog. https://cloud.google.com/blog/products/data-analytics/how-the-open-knowledge-format-can-improve-data-sharing

[2] MindStudio. (2026, July 3). What Is the Open Knowledge Format (OKF)? Google's Standard for AI Knowledge Bases. https://www.mindstudio.ai/blog/what-is-open-knowledge-format-okf-google-ai-knowledge-bases