McKinsey says ‘about half’ of its employees are using generative AI

McKinsey and Company, the global consulting firm with more than 30,000 employees across 67 countries, is embracing new generative AI tools in a major way: nearly 50 percent of the firm’s workforce is using ChatGPT and similar technology.

“About half of [our employees] are using those services with McKinsey’s permission,” said Ben Ellencweig, senior partner and global leader of QuantumBlack, the firm’s artificial intelligence consulting arm, during a media event at McKinsey’s New York Experience Studio on Tuesday.

Ellencweig emphasized that McKinsey had guard-rails for employees using generative AI, including “guidelines and principles” about what information the workers could input into these services.

“We do not upload confidential information,” Ellencweig said.

As for which exact AI services McKinsey employees are using, and for what purposes, the speakers remained tight-lipped.

However, another speaker at the event, Alex Singla, also a senior partner and global leader of QuantumBlack, implied that McKinsey was testing most of the leading generative AI services: “For all the major players, our tech folks have them all in a sandbox, [and are] playing with them every day,” he said.

Ellencweig and Singla were joined in Tuesday’s panel discussion on AI by Jacky Wright, another McKinsey senior partner and the firm’s chief technology and platform officer. The discussion was moderated by Ryan Heath, global tech correspondent from Axios. Other journalists in attendance among the dozens present at the event included representatives from The Wall Street Journal, CNBC and other leading media outlets.

The panelists related anecdotes about their own experiences with generative AI tools, as well as those of clients, including cautionary tales.

Singla described how one client, whose name was not disclosed, was in the business of mergers and acquisitions (M&A). Employees there were using ChatGPT and asking it “what would you think if company X bought company Y?,” and using the resulting answers to try and game-out the impact of potential acquisitions on their combined business.

“You don’t want to be doing that with a publicly accessible model,” Singla said, though he did not elaborate as to the reason why not.

The four Cs: how McKinsey customers are using gen AI

Ellencweig offered examples of how McKinsey clients and the businesses it researchers are currently using generative AI, which he called “The four Cs.” These include:

1. Coding: Ellencweig said some of McKinsey’s client software developers had seen productivity gains of 35-55% by using ChatGPT and similar tools.
2. Customer engagement: Some companies are using generative AI to offer more personalized customer interactions.
3. Creative content generation: Marketing firms are already using generative AI to streamline their content generation processes and to refine their audience segments, approaching a “segment of one,” i.e., marketing personalized for every individual.
4. Content synthesis: Firms are using generative AI to combine different data points and services in new ways.

McKinsey’s recommended 5-step approach for enterprise gen AI

Regarding companies where leaders are still wondering how to approach generative AI in a safe, secure, and smart way, Singla suggested they employ a five-step framework.

1. IT stack and infrastructure: “Before the model is built and you create these cool insights, you need to think about your IT stack, and infrastructure” and where the AI tools and data will be located — “in the cloud or your own infrastructure?”
2. Data: will you be using structured or unstructured data? Are you going to use your own data, proprietary data, third-party data, or some combination thereof? How will you organize this data? What protections do each of these require?
3. Choosing the right AI model: which LLMs or generative AI tools will your company deploy and why? Deciding this “is absolutely required, but not sufficient on its own,” said Singla.
4. UI and UX: Singla cited that ChatGPT’s simple interface as key to its uptake. “Anybody can use it, whether you’re eight years old or 80.”
5. Change management: how can the organization ensure that those using AI will be supported, have their questions answered, and their work duties changed by AI?