PwC and AWS Forge Path for Regulated AI Adoption

The integration of generative AI in various industries is on the rise, but with it come challenges, especially regarding accuracy and compliance. Sectors such as healthcare, financial services and pharmaceuticals are particularly under scrutiny to ensure that the AI systems do not produce inaccurate or misleading information. This issue is critical as companies work to instil robust controls and verification systems to manage AI outputs and address the phenomenon known as 'hallucinations'—instances where AI models generate plausible yet incorrect data.

Overcoming AI challenges with data chunking and reliability assurance guides (RAG) has therefore become a focal point in the adoption of enterprise AI, prompting organisations to explore effective implementation strategies that adhere to regulatory compliance.

The professional services giant PwC has broadened its collaboration with Amazon Web Services (AWS). This partnership, unveiled at AWS re:Invent 2024, centres on deploying foundational systems offered by Amazon's Bedrock—a platform designed to support the construction of Gen AI applications with a strong emphasis on text generation models that mirror human-like writing.

New checks to bolster AI reliability

The core of this collaboration is integrating new automated reasoning checks within Amazon Bedrock. These are designed to prevent the AI 'hallucinations' by applying logical verification steps to the outputs.

The verification system utilises automated reasoning to confirm the accuracy of AI-generated content and provides documentation on how its conclusions are derived. This development is particularly significant in regulated industries, where maintaining high accuracy is non-negotiable.

Matt Wood, Global Commercial Technology & Innovation Officer at PwC, emphasises the significance of these advancements: "In a field where breakthroughs are happening at incredible speed, reasoning is one of the most important technical advances to help our joint customers succeed in Gen AI."

These systems will be rolled out across PwC's product suite, including its Maestro Medical Legal Review solutions that aid in the analysis of healthcare documentation, ensuring all AI-generated content adheres to stringent industry regulations and company protocols, even in complex areas such as call centre operations.

Focus on industry-specific AI solutions

The prolonged collaboration between PwC and AWS isn't new but continues to evolve. Previous joint efforts include the development of AI annotation systems used by clients like Amazon for processing entertainment rights agreements and capable of digitising a range of documents to extract critical business data.

Kathrin Renz, Vice President of Industries at AWS, underlines the strategy: “Model selection and security controls are fundamental for unlocking the transformative potential of Gen AI in enterprise applications.”

Recently, PwC's US 2024 Cloud and AI Business Survey revealed that organisations documenting productivity gains from generative AI find marked benefit in solutions tailored to specific industries over generic applications. This partnership intends to cater to this need by developing bespoke AI solutions for highly regulated sectors.

Additionally, these sector-specific applications will be supported by Amazon's Nova foundation models, which will provide enterprise-grade intelligence for a variety of tasks through Amazon Bedrock and its new automated reasoning capabilities.

Companies like Continental Resources and WellSense have already engaged with PwC and AWS to finesse their generative AI strategies, leveraging both the deep industry knowledge of PwC and the technical prowess of AWS to fast-track their AI ambitions while ensuring compliance and accuracy are kept front and centre.

Explore the latest edition of Technology Magazine and be part of the conversation at our global conference series, Tech & AI LIVE.

Discover all our upcoming events and secure your tickets today.

Technology Magazine is a BizClik brand