As enterprises worldwide grapple with implementing Gen AI responsibly, the strengthened collaboration between IBM and AWS represents a pivotal moment in enterprise AI adoption.

This partnership arrives amid growing concerns about AI governance and security, particularly as organisations face increasing regulatory scrutiny and data protection challenges.

Against a backdrop where according to McKinsey, 65% of global enterprises plan to increase their AI investments in 2024, IBM’s integration of its Granite models and watsonx.governance with AWS’s infrastructure addresses critical gaps in enterprise-grade AI deployment.

The timing is especially relevant as markets in Asia-Pacific and Europe accelerate their AI initiatives while navigating complex regulatory frameworks like the EU AI Act.

The strategic IBM and AWS partnership

IBM and AWS have been working together to reshape how organisations manage their data and enhance AI governance.

This partnership is particularly significant as businesses increasingly rely on AI technologies to drive innovation and efficiency.

According to Chris Grusz, Managing Director of Technology Partnerships at AWS: “IBM is one of our largest partners for both technology and consulting and together we are at the forefront of enabling the democratisation of AI.”

This collaboration aims to facilitate the responsible deployment of AI solutions that meet the specific needs of enterprises.

Granite models

At the core of IBM’s Gen AI strategy is its Granite series of models, which are designed to help customers maximise the value of their data.

These models are built on open-source technology and are integrated into AWS environments, including Amazon SageMaker and Amazon Bedrock.

This integration allows organisations to optimise their AI initiatives while ensuring compliance with governance standards.

IBM has also focused on delivering high-performing models tailored for business applications.

The company is working closely with its business partners to support clients from initial pilot projects through to full production, thereby providing added value while distributing IBM software via the AWS Marketplace.

Driving responsible AI with IBM granite models

As enterprises expand their use of AI, selecting the appropriate model becomes crucial.

IBM Granite series

The IBM Granite series consists of open, enterprise-grade foundation models that empower AWS customers to scale their AI initiatives effectively.

These models are now available on Amazon Bedrock and Amazon SageMaker JumpStart, providing users with trusted, high-performance options tailored to specific business needs.

Granite 3.0 

The latest Granite 3.0 models are also now accessible on Amazon SageMaker JumpStart, allowing enterprises to implement advanced AI capabilities quickly.

AWS Neuron tech

To enhance flexibility for customers regarding computing resources, Granite models are also supported by AWS Neuron technology.

This allows deep learning workloads to run efficiently on AWS Inferentia and AWS Trainium instances.

By integrating with AWS Neuron technology, customers can run deep learning workloads efficiently on purpose-built hardware.

Trainium2

Looking ahead, IBM is exploring support for Trainium2, the latest generation of AWS AI chips designed for high-performance training and inference tasks.

The partnership also emphasises governance as a key component in deploying AI responsibly.

IBM watsonx.governance

IBM watsonx.governance provides a comprehensive framework for overseeing the entire AI model lifecycle.

In December, IBM and AWS plan to launch a streamlined user experience that integrates watsonx.governance directly with SageMaker.

This will enable users to conduct risk assessments and manage model approvals more effectively.

Embedding governance, security and observability into the AI lifecycle

The focus on security and trust is paramount in discussions about AI today.

In collaboration with AWS, IBM has positioned governance as an essential enabler of AI projects.

By providing insights and explainability, organisations can scale their initiatives confidently on platforms like Amazon SageMaker and Amazon Bedrock.

The partnership is also ensuring the automation portfolio is optimised to help organisations prioritise, streamline and increase operational efficiency to ensure AI projects deliver value, including:

Enhanced IBM watsonx.governance for AWS SageMaker: This release integrates watsonx.governance console directly with SageMaker, offering customisable risk assessment and streamlined model approval workflows.

The new functionality enables direct sharing of model data from SageMaker Model Registry, creating a unified AI governance framework.

IBM Guardium AI security available on AWS: This new SaaS offering helps organisations discover AI deployments (including shadow AI), ensure compliance, address vulnerabilities and protect sensitive data.
It integrates with watsonx.governance for comprehensive AI monitoring and security management.

IBM Instana Gen AI observability preview: This delivers advanced monitoring for Amazon Bedrock and SageMaker deployments.
It features Turbonomic-powered automated resource optimisation to improve performance whilst reducing cloud costs and includes expanded support for Amazon EKS and ECS with automated full-stack observability.

IBM concert enhancement preview: This new Gen AI-powered capability launching early 2025 improves application resilience across AWS infrastructure.
It identifies operational risks, predicts potential issues and recommends remediation steps to enhance platform stability and best practices adoption.

IBM autonomous security for cloud launch: This release introduces AI-driven security automation leveraging Amazon Bedrock.
It is designed to streamline cloud security management and decision-making processes for organisations scaling their AWS deployments.

“IBM has gone far beyond making their technology available on our platforms, they are prioritizing integration with AWS at each layer of the AI stack, making it easier for our mutual customers to choose the technology that best fits their needs and benefit from responsible AI in new ways”, Chris concludes.

https://technologymagazine.com/articles/how-ibm-aws-partnership-is-advancing-responsible-gen-ai