PwC to provide assurance for AI systems to clients

PwC has announced the launch of ‘Assurance for AI’ for clients in the UK. The set of services will aim provide independent assurance and other related solutions over AI systems.  

Since the public roll-out of OpenAI’s ChatGPT model in late 2022, the biggest players in the professional services sector have rushed to hitch their wagon to the booming artificial intelligence sector. For all the hypothetical power of AI, on a practical basis, it has yet to prove its value in many business cases.

There is no killer app, beyond an endless wave of re-skinned chatbots, and even when they do move to adopt it, many firms struggle to find ways they could enjoy returns on the investment – meaning most projects are trapped in the experimentation phase. On top of everything else, this is not helped by lingering trust issues in a technology that is widely acknowledged as being prone to ‘hallucinations’ and privacy and intellectual property violations.

Big Four firm PwC has announced a new offering, which might help clients to move beyond those last concerns, at least. The launch of the firm’s formal independent assurance services over relevant AI controls is aimed at helping businesses to meet this growing need to build trust in their AI systems.

The new offering complements, but is distinct from, PwC’s existing advisory services around managing AI risks, controls and governance, and testing and validating AI tools. Instead, the assurance line will provide confidence that they are designed, deployed, and operated responsibly, transparently, and aligned with regulatory expectations. By demonstrating this accountability and trustworthiness, PwC says “organisations can strengthen their position and credibility in the market”. 

Marc Bena, UK audit chief technology officer at PwC, commented, “AI is increasingly becoming core to business innovation and decision-making and client demand is growing. To get the most from AI, organisations need to trust the solutions, the data behind them, and the decisions they drive. Companies are developing proofs of concepts for new AI products but are hesitant about ‘pressing go’ and moving to production. Independent assurance can give organisations the additional confidence they need to take that next step.”

Artificial assurance

PwC has been providing advisory services for years around the deployment of Responsible AI, testing and validation services.  These services include assessing governance frameworks and control environments, and a analysing insights into how company procedures compare to leading practices and emerging standards and regulations. 

To meet the needs of an evolving risk, regulatory and business landscape, PwC will now provide its ‘Assurance for AI’ suite of services across the UK, delivered by multidisciplinary teams that combine technical knowledge of AI including machine learning, generative AI and natural language processing with proven expertise in audit, risk management, internal controls, attestation services, and external standards. 

Leigh Bates, PwC UK and global AI risk leader, commented, “AI is transforming business at every level, bringing unprecedented opportunities, but also a critical need for trust. At PwC, we’re leading the next evolution of AI assurance services, combining our deep technical insight with our global assurance expertise to help organisations navigate the complexities of AI responsibly and safely. By championing transparency, fairness, and accountability, we empower our clients to build credible, trustworthy systems and adopt AI with clarity, confidence, and purpose.” 

Leveraging their reputations as experts in audit, the other Big Four firms of  EY, Deloitte and KPMG are also said to be rushing to develop AI assurance services – letting their clients know which AI products are trustworthy, and which are not. Richard Tedder, an audit partner at Deloitte, recently told The Financial Times that AI assurance as “critical” for widespread AI adoption. This is especially the case for businesses relying on AI for crucial functions, and individual consumers using AI for personal matters such as health or finance.