Industry titans are creating their own internal AI bots to keep sensitive data secure and avoid the risks associated with public chatbots like ChatGPT. Many companies have restricted employee use of widely available AI tools, especially when handling confidential client information.
AI Adoption Across Sectors
In consulting, McKinsey’s internal AI platform, Lilli, is used monthly by about 75% of its 40,000 global employees to assist with junior-level tasks such as research and slide preparation. While automation hasn’t necessarily led to fewer jobs, McKinsey has cut roughly 5,000 positions since Lilli’s 2023 launch.
In banking, Morgan Stanley’s AI tools have saved coders around 280,000 hours this year by modernizing legacy code. Goldman Sachs’ AI allows employees to generate 95% of an IPO prospectus in minutes, a task that once took a six-person team over two weeks.
Retailers have also embraced AI: Walmart’s Trend-to-Product tool reduced its design timeline from six months to six to eight weeks by generating mood boards from internet trends. Target and Amazon introduced custom chatbots last year to replace ChatGPT for internal use.
Not all AI deployments have gone smoothly. UnitedHealth Group’s Optum healthcare unit faced a security incident in December when its internal AI chatbot, used for claims advice, was mistakenly made publicly accessible. The company said the tool was a demo that was never widely used and quickly restricted access.
What The Author Thinks
Companies building private AI systems reflect the need to control sensitive data and tailor tools to specific business needs. However, incidents like the Optum chatbot’s public exposure remind us that security is paramount. Private AI can boost efficiency and protect privacy, but only if implemented with rigorous oversight and transparency. Businesses must balance innovation with responsibility to ensure AI delivers real value without compromising trust.
