As companies race to adopt artificial intelligence, many are finding the most difficult part isn’t creating models – it’s making sure that data entering them is reliable and compliant with regulations. Without proper AI governance in place, businesses risk making flawed decisions, breaching privacy regulations or worse.
Collibra Enterprise Data Intelligence Services announced today a suite of AI tools intended to address this challenge with governance, automation and democratization capabilities that aim to create trust within AI systems.
“Our customers are eager to deploy AI use cases successfully, yet are coming to appreciate that there is more involved than simply applying algorithms,” explained Collibra CEO Felix Van de Maele in an interview with VentureBeat. As data is central to AI systems, strong data governance practices should also be prioritized as an essential component.
AI Governance Brings Order to Enterprise AI A key new offering from enterprise AI vendors is their AI Governance application, which serves as a command center for scattered AI initiatives across an enterprise. The tool seeks to bridge the divide between data scientists prototyping models and compliance/risk stakeholders who must approve them prior to production.
“We assist organizations in providing trusted AI by helping them better govern the AI use cases,” stated Van de Maele. “There can be something called’shadow AI,’ in which organizations experiment and explore without having any visibility over what’s happening. A major challenge is making sure data scientists or engineers find appropriate data which they are allowed to utilize while remaining compliant.”
“To bring AI use cases from concept to production successfully,” he stated, “you need change management systems in place as well as processes,” with multiple stakeholders including data scientists as well as legal, risk, compliance issues coming together as there are already regulations in place and more are to come.
Collibra has developed its data governance platform over a decade. AI Governance sits atop it; Van de Maele sees AI governance as an extension of data governance with many fundamental principles being the same between them.
With capabilities for defining policies, roles and responsibilities, this tool seeks to offer a standard workflow for registering, approving, documenting and monitoring AI use cases across an organization. The aim is to give stakeholders visibility into how AI is being applied while also giving them confidence that its use is taking place responsibly.
Collibra AI automates curation and stewardship tasks
Also included with this release is Collibra AI, which leverages large language models (LLMs) to automate many tedious data management tasks that had previously required manual effort.
Collibra AI’s automated metadata generation feature can provide valuable help in data cataloging and governance. When viewing complex database tables or column names without context, users often lack any idea what the information means; Collibra AI can generate the metadata by analyzing values for you.
Van de Maele explained, “it’s kind of a copilot approach,” wherein the system generates draft documents which users review and approve for final submission to their data sources. These details and definitions provide trustworthiness to ensure data accuracy and accessibility.
