Akeneo Agentic Product Data Release Puts AI Agents Under Control

Akeneo Agentic Product Data is the theme of the company's latest release, which has given its AI agents a much bigger job, and a much tighter leash at the same time. The Boston company announced its Autumn Release 2026 on October 28, 2026, a package that hands agents more of the daily work of product data while keeping every action inside enterprise permissions, traceable changes, and human approvals. For retailers and manufacturers whose catalogs now feed AI shopping assistants as well as human buyers, the release treats governed product data as the foundation of agentic commerce.

According to the company, the release builds on the agentic capabilities introduced for Ziggy, its agentic orchestrator, in the summer. Where that earlier step brought agents into product data work, the autumn release extends them across Shopify storefronts, supplier data, and product modeling. The through line is control: agents may enrich thousands of products and publish to many channels at once, but only inside what Akeneo calls a domain-specific AI harness, with the permissions, context, and orchestration that let AI act reliably inside a system of record rather than simply generate content.

Agents take on product work from source to storefront

Ziggy, the agentic orchestrator at the center of the release, now works with Shopify to help teams prepare and publish product data to their storefronts. The Akeneo Shopify connector now comes in two editions, Standard for deployments of up to three stores and Enterprise for organizations running many stores, according to the announcement. AI-assisted Product Model Suggestions help organizations bring order to large catalogs by automatically suggesting groupings of products, giving teams structure and repeatability as catalogs grow.

Supplier data gets a similar treatment. Akeneo Supplier Data Manager gains Supplier Collaboration and Supplier Catalog Management, capabilities designed to turn supplier contributions into structured data continuously rather than through periodic manual cleanup. That matters because supplier information is often where catalog quality breaks down first, and agents that can normalize it on arrival save merchandising teams long cycles of chasing and reformatting.

The release also looks outward to how products are discovered. As more product discovery happens through AI assistants and agents, the data those agents read becomes a competitive question. PX Insights now automates analysis of how products are discovered and how ready they are to be found by AI, feeding those signals back into the product data itself. In practice, that means a product page is no longer judged only by whether a human shopper finds it appealing, but by whether the agents doing the searching can read it, trust it, and recommend it. Related coverage of agent teams working under governance can be found in recent reporting on agentic observability for enterprise IT teams.

Control decides who acts, governance decides what ready means

Every new AI capability in the release runs inside the same AI harness behind Ziggy, according to the company. Ziggy works within the same role-based permissions as the human team, records every action as a traceable change, and proposes changes that a person approves before anything is applied. That approval step is the difference between an agent that drafts and an agent that publishes, and Akeneo has kept it firmly on the human side.

New Enterprise Access and Security Governance automates the user identity lifecycle with just-in-time provisioning and SCIM support. Authorized users get the right access through their corporate identity, and former employees lose it promptly. SIEM connectivity brings product information activity into the enterprise security operations picture, with audit trails for compliance work. The same control extends to supplier ecosystems through Supplier Enterprise Security, which brings single sign-on and multi-factor authentication to Supplier Data Manager, while transparent AI labeling identifies AI-generated content so teams and channels can see what an agent wrote and what a person validated.

Governance gets its own headline feature in Readiness Policies. Teams can define precisely what ready means for each channel, market, or business objective by combining rules on values, categories, descriptions, and relationships. Each product earns a readiness score, letting teams manage a portfolio and prioritize effort where it moves the needle. Product Compatibility adds fitment rules that consistently link products, spare parts, and accessories, so new products become compatible automatically as the catalog evolves instead of through spreadsheet wrangling. Compatibility data is then available through the product information system and its APIs to every downstream system.

An acquisition and a signal about agentic commerce

Alongside the release, Akeneo announced the acquisition of Dataggo, a long-standing AI-native partner specialized in product information integration and services for enterprise customers. According to the company, the deal strengthens its ability to deliver complex enterprise deployments and shorten time-to-value, bringing in deep experience connecting its platform to the systems around it. The release also introduces Private Showrooms in Akeneo Digital Showroom and improves the event platform, and it is available starting October 28, 2026, with specific capabilities varying by Akeneo Product Cloud edition.

The broader signal is one the whole agentic commerce wave keeps repeating: the agent economy will run on data that agents can trust, and Akeneo Agentic Product Data is built for exactly that test. Marketplaces are already learning the same lesson, as recent reporting on trusted agentic selling showed, with buyers and sellers asking for transparency and a final human say. Akeneo is betting that the same trust requirement applies inside the catalog itself. Agents that enrich, publish, and recommend at machine scale are only useful if every change is attributable, every rule is explicit, and a person can still say no. The Autumn Release 2026 is, in that sense, less a showcase of what agents can do than a statement about the conditions under which enterprises will let them do it. For more detail, see the announcement coverage from The Agile Brand Guide and Akeneo's own site.