PricewaterhouseCoopers to Pour $1 Billion Into Generative AI

Generative AI tools are designed to generate natural-language responses, images or computer code from user text prompts. 

For PwC, the goal isn’t only to develop and embed generative AI into its own technology stack and client-services platforms, but also advising other companies on how best to use generative AI, while helping them build those tools, said

Mohamed Kande,

PwC’s vice chair and co-leader of U.S. consulting solutions and global advisory leader.

Mr. Kande said the company would pay to access OpenAI’s GPT-4 language model, the underlying software that drives ChatGPT, to build and run apps in Microsoft’s Azure cloud. While ChatGPT is a free online tool, OpenAI charges developers to access its language model and create their own software tools. The model, which was recently upgraded, is trained on massive stores of language data gathered from online posts, interviews and other sources, to understand natural-language prompts and produce intelligible responses.  

Once the models are fully trained and tested, Mr. Kande sees the technology being used to quickly write reports and prepare compliance documents, analyze and evaluate business strategies, identify inefficiencies in operations or create marketing materials and sales campaigns, among many other applications.

“This is about using generative AI to run the company in a more efficient way,” Mr. Kande said. “Embracing this technology is critical.”

Eric Boyd,

corporate vice president of Microsoft’s AI platform, said the move will enable PwC to access OpenAI’s generative AI tools with the added compliance and data security of its Azure cloud-computing service. Microsoft itself announced a multiyear, multibillion-dollar investment in OpenAI, a startup launched in 2015. 

Mr. Boyd said more than 1,000 organizations—including startups and multinational corporations—are now using OpenAI tools in Microsoft’s cloud in areas like customer support, conversational AI, summarization, writing assistance and customization by “gaining insights from data using search, data extraction and classification,” he said.

Other large accounting firms, including KPMG LLP and Ernst & Young, are also investing in generative AI. TurboTax owner

Intuit Inc.,

for instance, is building its own generative AI language model for financial management, trained on years of interactions with its business customers, the company said.

Accounting, tax preparation, auditing and other financial services are ripe areas for generative AI, analysts said. 

“Generative AI offers many attractive use cases for companies like this,” said

Rowan Curran,

an analyst at information-technology research firm

Forrester Research Inc.,

covering data science, machine learning, artificial intelligence and computer vision. 

Mr. Curran said large language models like GPT-4 can be used to help with information discovery and retrieval—particularly in exploring unstructured and semi-structured data—along with the potential to improve the process of preparing reports with a “much lower effort from the human auditors.” After the initial burst of interest in generative AI sparked by ChatGPT, which launched in November, companies are now moving into a phase of experimenting, building and deploying their first-generation applications, he said.

In a survey of about 500 corporate IT decision makers conducted by market research firm Enterprise Technology Research, 53% said they planned to evaluate, use or allocate further resources to OpenAI’s ChatGPT technology—a record for any single technology provider, said ETR’s Chief Strategist

Erik Bradley.

Consulting and business-services firms, along with educational institutions, were the leading sectors in plans to evaluate and use generative AI and large language models, according to the survey.

“The outsized investments in generative AI make perfect sense for organizations trying to do more with less,” Mr. Bradley said. “The real question is, will all of this initial investment and evaluation turn into actual business utilizations.”

Across all sectors, spending in the global generative AI market is expected to reach $42.6 billion by the end of the year, growing at a compound annual rate of 32% to $98.1 billion by 2026, according to market analytics firm PitchBook Data Inc.

While acknowledging the benefits,

Mark D. McDonald,

senior director analyst at IT research and consulting firm

Gartner Inc.,

said the use of generative AI in areas like tax preparation requires validation by a professional. It might create murky compliance issues, he said. “Referencing an algorithm as the rationale for tax decisions is not an excuse that auditors will accept,” Mr. McDonald said.

Mr. Kande said PwC isn’t aiming to replace workers with generative AI, but rather optimize their jobs by automating time-consuming, repetitive tasks. “We have 65,000 people in the U.S.,” he said about the firm’s U.S. operations. “We are not going to leave anybody behind. It’s going to be a team sport.”

Write to Angus Loten at [email protected]

Generative AI tools are designed to generate natural-language responses, images or computer code from user text prompts. 

For PwC, the goal isn’t only to develop and embed generative AI into its own technology stack and client-services platforms, but also advising other companies on how best to use generative AI, while helping them build those tools, said

Mohamed Kande,

PwC’s vice chair and co-leader of U.S. consulting solutions and global advisory leader.

Mr. Kande said the company would pay to access OpenAI’s GPT-4 language model, the underlying software that drives ChatGPT, to build and run apps in Microsoft’s Azure cloud. While ChatGPT is a free online tool, OpenAI charges developers to access its language model and create their own software tools. The model, which was recently upgraded, is trained on massive stores of language data gathered from online posts, interviews and other sources, to understand natural-language prompts and produce intelligible responses.  

Once the models are fully trained and tested, Mr. Kande sees the technology being used to quickly write reports and prepare compliance documents, analyze and evaluate business strategies, identify inefficiencies in operations or create marketing materials and sales campaigns, among many other applications.

“This is about using generative AI to run the company in a more efficient way,” Mr. Kande said. “Embracing this technology is critical.”

Eric Boyd,

corporate vice president of Microsoft’s AI platform, said the move will enable PwC to access OpenAI’s generative AI tools with the added compliance and data security of its Azure cloud-computing service. Microsoft itself announced a multiyear, multibillion-dollar investment in OpenAI, a startup launched in 2015. 

Mr. Boyd said more than 1,000 organizations—including startups and multinational corporations—are now using OpenAI tools in Microsoft’s cloud in areas like customer support, conversational AI, summarization, writing assistance and customization by “gaining insights from data using search, data extraction and classification,” he said.

Other large accounting firms, including KPMG LLP and Ernst & Young, are also investing in generative AI. TurboTax owner

Intuit Inc.,

for instance, is building its own generative AI language model for financial management, trained on years of interactions with its business customers, the company said.

Accounting, tax preparation, auditing and other financial services are ripe areas for generative AI, analysts said. 

“Generative AI offers many attractive use cases for companies like this,” said

Rowan Curran,

an analyst at information-technology research firm

Forrester Research Inc.,

covering data science, machine learning, artificial intelligence and computer vision. 

Mr. Curran said large language models like GPT-4 can be used to help with information discovery and retrieval—particularly in exploring unstructured and semi-structured data—along with the potential to improve the process of preparing reports with a “much lower effort from the human auditors.” After the initial burst of interest in generative AI sparked by ChatGPT, which launched in November, companies are now moving into a phase of experimenting, building and deploying their first-generation applications, he said.

In a survey of about 500 corporate IT decision makers conducted by market research firm Enterprise Technology Research, 53% said they planned to evaluate, use or allocate further resources to OpenAI’s ChatGPT technology—a record for any single technology provider, said ETR’s Chief Strategist

Erik Bradley.

Consulting and business-services firms, along with educational institutions, were the leading sectors in plans to evaluate and use generative AI and large language models, according to the survey.

“The outsized investments in generative AI make perfect sense for organizations trying to do more with less,” Mr. Bradley said. “The real question is, will all of this initial investment and evaluation turn into actual business utilizations.”

Across all sectors, spending in the global generative AI market is expected to reach $42.6 billion by the end of the year, growing at a compound annual rate of 32% to $98.1 billion by 2026, according to market analytics firm PitchBook Data Inc.

While acknowledging the benefits,

Mark D. McDonald,

senior director analyst at IT research and consulting firm

Gartner Inc.,

said the use of generative AI in areas like tax preparation requires validation by a professional. It might create murky compliance issues, he said. “Referencing an algorithm as the rationale for tax decisions is not an excuse that auditors will accept,” Mr. McDonald said.

Mr. Kande said PwC isn’t aiming to replace workers with generative AI, but rather optimize their jobs by automating time-consuming, repetitive tasks. “We have 65,000 people in the U.S.,” he said about the firm’s U.S. operations. “We are not going to leave anybody behind. It’s going to be a team sport.”

Write to Angus Loten at [email protected]