Photo credit: venturebeat.com
Stay updated on the latest in AI with our daily and weekly newsletters.
The AI governance firm Credo AI has introduced a new platform aimed at enhancing oversight of responsible AI practices by integrating with various third-party AI operations and business tools.
Named the Integrations Hub, this recently launched platform allows businesses to connect their generative AI development tools—such as Amazon SageMaker, MLFlow, and Microsoft Dynamics 365—to a centralized governance system. Additionally, it supports integration with deployment platforms like Asana, ServiceNow, and Jira.
The primary purpose of the Integrations Hub is to facilitate organizations working on AI applications in linking their efforts to Credo AI’s governance framework. Instead of submitting documentation that demonstrates compliance with safety and security standards, the hub automatically collects relevant metadata from those applications.
According to Credo AI, the Integrations Hub will simplify compliance by connecting directly with existing model stores, ensuring that data for regulatory checks is readily uploaded to the governance platform. It will also gather necessary datasets for comprehensive governance oversight.
Navrina Singh, the founder and CEO of Credo AI, explained to VentureBeat that the hub was developed to embed AI governance into the development lifecycle from the onset. “Organizations, especially those in the Global 2000 category, are rapidly adopting AI technologies and exploring new AI tools,” Singh remarked. “Our goal was to empower these companies to maximize their AI investments while making governance straightforward and accessible.”
The Integrations Hub comes equipped with predefined connections to notable platforms, including Jira, ServiceNow, Amazon SageMaker and Bedrock, Salesforce, MLFlow, Asana, Databricks, Microsoft Dynamics 365, Azure Machine Learning, Weights & Biases, Hugging Face, and Collibra. Additional integrations can be customized for a fee.
Embedding Governance Early
As businesses seek to navigate the complexities surrounding generative AI responsibly, new opportunities have emerged for companies to address risk management and compliance challenges. Credo AI is among several firms striving to streamline responsible AI practices for organizations.
IBM’s Watsonx offers a governance platform that evaluates models based on accuracy, bias, and compliance, while Collibra has introduced a comprehensive suite of AI governance tools designed to facilitate documentation and monitoring of AI initiatives.
While Credo AI evaluates applications for potential brand risks such as accuracy, its main focus is to help organizations adhere to established regulations regarding automated systems and any anticipated future regulations. Although there are currently few formal regulations governing generative AI, existing data privacy and retention policies are already in place for many enterprises thanks to the earlier compliance efforts in machine learning and data governance.
Singh highlighted that certain regions mandate AI governance reporting from enterprises. She referred specifically to New York City Law 144, which prohibits the use of automated tools in employment decision-making. “Specific metrics, like the demographic parity ratio, must be documented,” Singh explained. “Credo AI aligns with this law by integrating it into our AI operations system, enabling us to capture the necessary metadata to fulfill legal obligations.”
Source
venturebeat.com