{"version":"1.0","type":"agent_native_article","locale":"en","slug":"databricks-bets-ontology-controls-brain-enterprise-ai-agents-mqkkyu5g","title":"Databricks Bets on Ontology and Reveals Who Controls the Brain of Enterprise AI Agents","primary_category":"innovation","author":{"name":"Lucía Navarro","slug":"lucia-navarro"},"published_at":"2026-06-19T06:03:43.039Z","total_votes":82,"comment_count":0,"has_map":true,"urls":{"human":"https://sustainabl.net/en/articulo/databricks-bets-ontology-controls-brain-enterprise-ai-agents-mqkkyu5g","agent":"https://sustainabl.net/agent-native/en/articulo/databricks-bets-ontology-controls-brain-enterprise-ai-agents-mqkkyu5g"},"summary":{"one_line":"Databricks launches Genie Ontology, a living semantic layer that gives AI agents a single authoritative source of business definitions, escalating the race to own the enterprise AI control plane.","core_question":"Who will control the semantic infrastructure that enterprise AI agents rely on to reason and act — and what does it take to make that control durable?","main_thesis":"Databricks is positioning Genie Ontology not as a standalone product but as the semantic foundation of an agentic system of record, betting that whoever owns business definitions owns the enterprise AI stack. The bet is strategically coherent but operationally contingent: its value depends on data maturity, governance discipline, and continuous maintenance that most organizations do not yet have."},"content_markdown":"## Databricks bets on ontology and reveals who controls the brain of enterprise AI agents\n\nThe history of enterprise artificial intelligence can be measured in layers. First came vector databases, which enabled semantic similarity searches across large volumes of text. Then came retrieval-augmented generation — RAG, as it is known by its acronym — which combined language models with external knowledge sources to reduce hallucinations. That architecture dominated the last two years and became the de facto standard for building corporate assistants.\n\nNow Databricks is betting that architecture is not enough. At its annual Data + AI Summit conference, CEO Ali Ghodsi presented **Genie Ontology**, a context layer that automatically extracts business definitions from internal data, dashboards, SQL queries, documents, pipelines, and applications, and organizes them into a living graph that AI agents can consult to understand how an organization operates. The product is in preview phase and uses a ranking system inspired by Google's PageRank to determine which source deserves the most authority: who created the information, how much it is used, whether it is linked to certified assets, and when it was last updated.\n\nThe move is not purely technical. It is a declaration of intent about who will control the semantic infrastructure of the future enterprise, and that dispute has first-order economic consequences.\n\n## From archive to authority\n\nThe problem that Genie Ontology attempts to solve is not new. In any medium-sized or large company, the definition of \"monthly recurring revenue\" can differ between finance, sales, and the data team. Three departments, three different numbers for the same metric. Traditional RAG systems do not solve that: they retrieve what appears similar to the question, but they do not distinguish between an official definition and one that someone wrote in a Google document three years ago.\n\nAn ontology, on the other hand, does not merely retrieve; **it encodes hierarchical relationships between concepts**, establishes which source holds authority over which definition, and allows different AI agents to share the same business vocabulary. Michael Leone, an analyst at Moor Insights & Strategy, describes it with clarity: a single definition feeding all agents means you stop receiving three different answers to the same question. The operational value of that consistency, in organizations where critical decisions are made based on automated reports, is high.\n\nAshish Chaturvedi, a researcher at HFS Research, goes further and links this to the most persistent obstacle to corporate AI adoption: the lack of trust. According to his analysis, the central problem is not technical but one of knowledge governance. Decision-makers do not act on AI outputs because they cannot trace where they come from or verify whether the reasoning chain used the correct sources. An ontology anchored in official definitions with traceability back to the source directly addresses that deficit.\n\nDatabricks also integrates Genie Ontology with its Unity Catalog Semantics platform, which allows organizations to upload their own definitions or corporate vocabularies and maintain control over what enters the graph. Internally, the company reports having generated around **4.5 million ontological fragments** during its own testing process. That gives an idea of the scale of the problem they are attempting to solve and, at the same time, of the complexity of keeping it up to date.\n\n## The risk that the narrative of progress omits\n\nEvery architecture has its limits. Stephanie Walter, of HyperFRAME Research, identifies the missing link with precision: **verification**. An ontology improves the context in which an agent operates, but it does not guarantee that the answer is correct. An agent can consult the correct definition and still apply flawed logic, omit rows in a dataset, misinterpret a workflow, or take an unintended action. Semantic consistency is not the same as operational correctness.\n\nThat distinction matters especially because the horizon Databricks is targeting is not query assistants but **agents that execute actions**: modifying pipelines, generating regulatory reports, triggering alerts, or making automated decisions in business processes. In that context, a well-grounded semantic error can be more dangerous than an obvious ambiguity, because it travels further before anyone detects it.\n\nLeone adds another dimension: most companies do not have the data maturity and governance that implementing an ontology layer with rigor requires. If data lineage is weak, metric owners are not defined, or the current definitions are contradictory, adding an ontology does not solve the problem — it accelerates it. The graph feeds on existing sources, and if those sources are inconsistent, the inconsistency propagates with greater speed and the appearance of authority.\n\nWalter adds the quietest dimension of risk: maintenance. An ontology is not a project that is configured once. It is a living asset that needs to be updated every time the business changes, every time a new product is launched, every time a metric is redefined or a unit is reorganized. Without update processes, clear ownership, and mechanisms for resolving conflicts between definitions, the graph becomes obsolete. And an obsolete ontology with algorithmic authority over agents is, according to Walter, \"another stalled metadata project with a more sophisticated name.\"\n\nThat does not invalidate Databricks' bet, but it does define the terrain on which the product will have to demonstrate its value: not in a presentation on a stage, but in the operational maintenance within organizations with imperfect data and governance structures that are still maturing.\n\n## The dispute over the enterprise control plane\n\nGenie Ontology does not exist in a vacuum. Snowflake has Horizon Context, its own semantic layer for agents. Microsoft is building equivalent capabilities within Copilot, Fabric, and its IQ family — Work IQ, Fabric IQ, Foundry IQ — integrating business context and governance into its broader infrastructure. The problem, Leone notes, is that each vendor has branded a basically similar idea with a different name, and that terminological fragmentation slows adoption because CIO teams cannot clearly compare what they are evaluating.\n\nBeyond the names, what is in dispute is structurally significant. Chaturvedi describes it as the race to become the **enterprise AI control plane**: the place where data, governance, semantics, and agent execution converge. The historical analogy he uses is precise: ERP systems became the system of record for business transactions; data warehouses became the system of record for analytics. Now the question being decided is which platform becomes the system of record for AI agents.\n\nDatabricks is positioning Genie Ontology within a broader architecture that includes LTAP — its proposed foundation for agentic applications — and OpenSharing, designed to reduce integration costs in corporate AI environments. Connected together, these components point toward a vision that Ghodsi himself describes as an \"agentic system of record\": an authoritative source from which agents read, reason, and act. It is not an isolated product; it is a platform strategy.\n\nThe structural advantage of data providers in this race is real: they already own the data, governance controls, lineage, and permissions that agents need to operate safely. That puts them in a different position from a model provider or an orchestration tooling vendor. But that advantage has a less favorable side: it also makes them dependent on their customers already having their data in order. And for most companies, that is still not the case.\n\nChaturvedi offers a heuristic that simplifies the decision for teams currently evaluating these options: the context layer follows the gravity of the data. If the data lives in Databricks, Genie Ontology is the natural path. If it is in Snowflake, Horizon Context is. If the infrastructure is predominantly Microsoft, the IQ family is the route. Bhupendra Chopra, from the consulting firm Kanerika, reinforces that argument: above the marketing of each platform, the real decision is made by where the data already resides.\n\nSnowflake is attempting to differentiate its offering by betting on open semantic interoperability, which in theory allows business definitions to move between platforms without becoming trapped in a single vendor's data model. That bet directly targets the risk of semantic lock-in — the equivalent of platform lock-in, but applied to the corporate vocabulary — in environments where companies operate across multiple data systems simultaneously.\n\n## Value is captured where execution is verified\n\nThe dominant narrative around these platforms speaks of context, consistency, and trust. All of those dimensions matter, but there is one that still has no solid answer in any of the available proposals: **how to verify that what the agent did was the right thing.**\n\nThat is the real frontier. Not the quality of the context with which the agent begins a task, but the ability to audit — with complete traceability — what the agent did, which definitions it used, which data it processed, what logic it applied, and whether the result is reproducible. Walter summarizes it without ambiguity: the next battleground in enterprise AI is not context, but verifiable execution.\n\nThat has direct consequences for where economic value is captured in this race. An ontology that improves semantic consistency is a valuable asset, but it is not sufficient for an organization to be able to delegate operational decisions with real consequences — financial, regulatory, operational — to autonomous agents. For that level of delegation to occur, the platform needs to offer something more: an auditable record of decisions, correction mechanisms for when the agent makes a mistake, and guarantees about what happens when the context changes and the graph has not yet been updated.\n\nDatabricks is building in that direction, although Genie Ontology alone does not yet answer that question. What the full set of announcements at the Data + AI Summit reveals is a coherent strategy toward that objective: data + governance + semantics + agentic execution as integrated layers within a single platform. The coherence of the vision is clear. The stress test will come when the ontology has to remain accurate within organizations that change faster than any graph can update itself.\n\nThat tension between the ambition of the architecture and the operational reality of the companies that will adopt it is where it will be decided whether this bet generates sustainable value — or whether it becomes sophisticated infrastructure built on foundations that are not yet ready to support it.","article_map":{"title":"Databricks Bets on Ontology and Reveals Who Controls the Brain of Enterprise AI Agents","entities":[{"name":"Databricks","type":"company","role_in_article":"Primary subject; developer of Genie Ontology and architect of the broader agentic platform strategy announced at Data + AI Summit."},{"name":"Ali Ghodsi","type":"person","role_in_article":"Databricks CEO who presented Genie Ontology and articulated the vision of an 'agentic system of record.'"},{"name":"Genie Ontology","type":"product","role_in_article":"The semantic context layer at the center of the article; the product whose architecture, value, and risks are analyzed."},{"name":"Unity Catalog Semantics","type":"product","role_in_article":"Databricks platform integrated with Genie Ontology for uploading and governing proprietary business definitions."},{"name":"Michael Leone","type":"person","role_in_article":"Analyst at Moor Insights & Strategy; provides perspective on the operational value of semantic consistency and data maturity prerequisites."},{"name":"Ashish Chaturvedi","type":"person","role_in_article":"Researcher at HFS Research; frames the adoption problem as knowledge governance and trust, and introduces the control plane race analogy."},{"name":"Stephanie Walter","type":"person","role_in_article":"Analyst at HyperFRAME Research; identifies the verification gap and maintenance burden as the critical unresolved risks."},{"name":"Bhupendra Chopra","type":"person","role_in_article":"Consultant at Kanerika; reinforces the data gravity heuristic for platform selection."},{"name":"Snowflake","type":"company","role_in_article":"Primary competitor; offers Horizon Context as its semantic layer and bets on open semantic interoperability to reduce lock-in."},{"name":"Microsoft","type":"company","role_in_article":"Competitor building equivalent semantic capabilities within Copilot, Fabric, and the IQ family (Work IQ, Fabric IQ, Foundry IQ)."},{"name":"Moor Insights & Strategy","type":"institution","role_in_article":"Research firm providing analyst commentary on semantic consistency and data maturity."},{"name":"HFS Research","type":"institution","role_in_article":"Research firm providing analysis on enterprise AI adoption barriers and the control plane race."}],"tradeoffs":["Semantic consistency vs. operational correctness: an ontology improves context but does not guarantee the agent's reasoning or actions are correct.","Speed of deployment vs. data maturity prerequisite: deploying an ontology on inconsistent data propagates inconsistency faster and with false authority.","Vendor integration depth vs. lock-in risk: deeper integration with a single platform's semantic layer increases switching costs as agent ecosystems grow.","Architectural ambition vs. maintenance burden: a living ontology requires continuous governance investment that most organizations are not yet structured to provide.","Centralized authority vs. distributed definition ownership: assigning algorithmic authority to a single source resolves ambiguity but may suppress legitimate definitional variation across business units.","Open interoperability vs. platform coherence: Snowflake's open semantic bet reduces lock-in but may sacrifice the tight integration that makes a closed platform more operationally reliable."],"key_claims":[{"claim":"Databricks' Genie Ontology uses a PageRank-inspired ranking to assign authority to business definitions based on creator, usage frequency, certification status, and recency.","confidence":"high","support_type":"reported_fact"},{"claim":"Databricks generated approximately 4.5 million ontological fragments during internal testing of Genie Ontology.","confidence":"high","support_type":"reported_fact"},{"claim":"Genie Ontology is integrated with Unity Catalog Semantics, allowing organizations to upload proprietary definitions and maintain control over graph content.","confidence":"high","support_type":"reported_fact"},{"claim":"The product was announced at the Data + AI Summit and is currently in preview phase.","confidence":"high","support_type":"reported_fact"},{"claim":"Traditional RAG systems cannot distinguish between authoritative definitions and informal ones, producing inconsistent answers across departments.","confidence":"high","support_type":"inference"},{"claim":"The primary adoption blocker for enterprise AI is knowledge governance and trust, not technical capability.","confidence":"medium","support_type":"inference"},{"claim":"An ontology built on inconsistent source data propagates inconsistency faster and with the appearance of authority, making it potentially more harmful than no ontology.","confidence":"medium","support_type":"inference"},{"claim":"Snowflake is differentiating by betting on open semantic interoperability to reduce semantic lock-in risk.","confidence":"medium","support_type":"reported_fact"}],"main_thesis":"Databricks is positioning Genie Ontology not as a standalone product but as the semantic foundation of an agentic system of record, betting that whoever owns business definitions owns the enterprise AI stack. The bet is strategically coherent but operationally contingent: its value depends on data maturity, governance discipline, and continuous maintenance that most organizations do not yet have.","core_question":"Who will control the semantic infrastructure that enterprise AI agents rely on to reason and act — and what does it take to make that control durable?","core_tensions":["Architectural ambition vs. organizational readiness: Genie Ontology's value is gated by data maturity and governance discipline that most target customers do not yet have.","Semantic authority vs. semantic correctness: encoding a definition as authoritative does not make the agent's use of that definition correct or its actions safe.","Living graph vs. maintenance reality: an ontology must update continuously as the business changes, but most organizations lack the processes and ownership structures to sustain that.","Vendor differentiation vs. customer comparability: each platform brands similar ideas differently, creating marketing noise that slows the adoption it is meant to accelerate.","Control plane consolidation vs. multi-cloud reality: the data gravity heuristic assumes data concentration, but most enterprises operate across multiple data systems simultaneously."],"open_questions":["How will Databricks handle ontological conflicts when the same metric is defined differently by certified sources within the same organization?","What update latency is acceptable before an ontology becomes operationally misleading for agents executing time-sensitive decisions?","Will open semantic interoperability (Snowflake's bet) become an industry standard, or will semantic lock-in follow the same trajectory as data lock-in?","How does Genie Ontology perform in organizations with weak data lineage and undefined metric ownership — the majority of its target market?","What does verifiable execution look like as a product, and which platform will credibly solve it first?","Can the PageRank-inspired authority ranking be gamed or manipulated by internal actors who control certified assets?","At what organizational data maturity threshold does deploying an ontology layer generate positive ROI vs. accelerating existing inconsistencies?"],"training_value":{"recommended_for":["CIOs and CDOs evaluating enterprise AI platform strategy","Data architects designing agentic application infrastructure","AI product managers assessing semantic layer vendors","Enterprise AI adoption consultants","Business analysts building the case for or against ontology investment","Researchers studying platform competition dynamics in enterprise software"],"when_this_article_is_useful":["When evaluating enterprise AI agent platforms and needing to compare Databricks, Snowflake, and Microsoft offerings on semantic layer capabilities.","When building a business case for or against deploying an ontology layer in an organization with incomplete data governance.","When advising on AI adoption strategy and needing to articulate why trust and traceability are the primary blockers, not technical capability.","When assessing vendor claims about semantic infrastructure and needing a framework to separate marketing terminology from structural differentiation.","When designing AI agent architectures that will execute consequential actions and need to understand the gap between context quality and execution auditability."],"what_a_business_agent_can_learn":["How to distinguish between semantic consistency (what an ontology provides) and operational correctness (what verifiable execution requires) when evaluating AI infrastructure.","How to apply the data gravity heuristic to simplify platform selection decisions: the context layer follows where the data already lives.","How to identify the hidden TCO of semantic infrastructure: maintenance burden, ownership assignment, and conflict-resolution processes are as important as initial deployment cost.","How to recognize when a vendor narrative omits the verification dimension — the gap between 'agent has good context' and 'agent did the right thing.'","How to assess organizational data maturity as a prerequisite before deploying governance-dependent AI infrastructure.","How the system of record succession pattern (ERP → data warehouse → AI agent semantic layer) predicts where durable platform lock-in will form in enterprise AI."]},"argument_outline":[{"label":"1. The RAG ceiling","point":"Retrieval-augmented generation became the de facto standard for corporate AI assistants but cannot distinguish between an authoritative definition and an informal one, producing inconsistent answers across departments.","why_it_matters":"This ceiling is the market gap Genie Ontology targets; understanding it clarifies why Databricks frames this as an architectural leap, not an incremental feature."},{"label":"2. What an ontology adds","point":"Unlike RAG, an ontology encodes hierarchical relationships between concepts, assigns authority to sources using a PageRank-inspired ranking, and gives all agents a shared business vocabulary.","why_it_matters":"Semantic consistency across agents reduces conflicting outputs and builds the trust that decision-makers need to act on AI-generated reports."},{"label":"3. The trust deficit in enterprise AI","point":"HFS Research analyst Ashish Chaturvedi identifies the core adoption blocker as knowledge governance, not technology: executives distrust AI outputs they cannot trace or verify.","why_it_matters":"An ontology with source traceability directly addresses this deficit, making it a governance tool as much as a technical one."},{"label":"4. The verification gap","point":"Stephanie Walter of HyperFRAME Research argues that semantic consistency is not operational correctness: an agent can consult the right definition and still apply flawed logic or take unintended actions.","why_it_matters":"For agents executing consequential actions — modifying pipelines, generating regulatory reports — a well-grounded semantic error is more dangerous than an obvious ambiguity because it travels further before detection."},{"label":"5. The data maturity prerequisite","point":"Michael Leone notes that most companies lack the data lineage, defined metric owners, and governance structures that a rigorous ontology layer requires. Adding an ontology to inconsistent data propagates inconsistency faster, with the appearance of authority.","why_it_matters":"This is the primary adoption risk: the product's value is gated by organizational readiness that Databricks cannot control."},{"label":"6. The maintenance burden","point":"An ontology is a living asset requiring continuous updates as the business changes. Without clear ownership and conflict-resolution processes, it becomes obsolete — and an obsolete ontology with algorithmic authority is, per Walter, 'another stalled metadata project with a more sophisticated name.'","why_it_matters":"Maintenance cost and governance overhead are the hidden TCO of semantic infrastructure, rarely surfaced in vendor narratives."}],"one_line_summary":"Databricks launches Genie Ontology, a living semantic layer that gives AI agents a single authoritative source of business definitions, escalating the race to own the enterprise AI control plane.","related_articles":[{"reason":"Directly examines the tension between AI agent autonomy claims and the governance structures required to make that autonomy safe — the same tension at the core of the verifiable execution problem Genie Ontology has not yet solved.","article_id":14001},{"reason":"Analyzes why 95% of enterprise AI projects fail to survive the pilot phase, providing the organizational readiness context that explains the adoption risks identified in the Genie Ontology analysis.","article_id":13655}],"business_patterns":["Platform gravity: the winning semantic layer will be the one co-located with the data, not the one with the best standalone architecture.","System of record succession: each major enterprise computing era produces a new system of record (ERP for transactions, data warehouse for analytics, now semantic layer for AI agents).","Governance as adoption gate: enterprise AI adoption is blocked more by trust and traceability deficits than by technical capability gaps.","Maintenance debt as hidden TCO: semantic infrastructure that is not continuously governed degrades into obsolete metadata with the appearance of authority.","Terminological fragmentation as adoption friction: when competing vendors brand similar capabilities with different names, CIO evaluation cycles slow and adoption lags.","Data maturity as prerequisite, not outcome: tools that assume governance maturity fail in organizations where that maturity is still being built."],"business_decisions":["Whether to adopt a semantic/ontology layer now or wait for the technology and organizational prerequisites to mature.","Which platform to build the enterprise AI context layer on, using data gravity (where data already resides) as the primary decision criterion.","How to assign ownership and update responsibilities for ontological definitions before deploying an ontology layer.","Whether to prioritize semantic consistency (what Genie Ontology offers) or verifiable execution (what no platform fully offers yet) in AI agent procurement criteria.","How to evaluate vendor claims about semantic infrastructure when terminological fragmentation (Genie Ontology vs. Horizon Context vs. IQ family) obscures meaningful comparison.","Whether to accept semantic lock-in risk with a single vendor or prioritize open semantic interoperability as Snowflake proposes."]}}