The rapid adoption of generative AI tools like OpenAI’s ChatGPT is transforming workplaces, but employees’ use of the technology often remains hidden due to a lack of clear corporate guidelines. According to the Financial Times, workers across industries are quietly experimenting with AI, filling gaps in company policies, and sometimes surpassing their more experienced colleagues in productivity.
AI in the Shadows
“Matt,” a pseudonymous 27-year-old researcher at a pharmaceuticals company, credits ChatGPT for helping him keep pace with his seasoned coworkers. Despite the tool’s contributions to his success, Matt has not disclosed his use of generative AI, citing embarrassment and uncertainty about his employer’s stance. “I didn’t want to admit to using shortcuts,” he told the Financial Times.
Matt is not alone. A survey conducted in August by the Federal Reserve Bank of St. Louis found that nearly 25% of U.S. workers regularly use generative AI, with adoption nearing 50% in industries like software and finance. Yet, many workers are forging their own paths due to the absence of formal policies.
Policies Lag Behind Adoption
By September, almost two years after ChatGPT’s launch, fewer than half of executives surveyed by U.S. law firm Littler had implemented rules for generative AI use. Those with policies often started with blanket bans, motivated by concerns over data privacy. Companies like Apple, Samsung, and Goldman Sachs initially prohibited employees from using external AI platforms but are now reconsidering such measures.
“We started at ‘block,’ but we didn’t want to maintain ‘block,’” Jerry Geisler, Walmart’s chief information security officer, told the Financial Times. Walmart encourages employees to use an in-house AI tool called My Assistant while monitoring external AI use on corporate devices. The company has adopted a “non-punitive” approach, aiming to guide employees toward safer AI practices rather than penalizing them.
AI Integration: A Strategic Shift
Organizations like McKinsey, JPMorgan Chase, and Linklaters are developing internal AI tools tailored for secure and efficient use. However, smaller companies with fewer resources face challenges in navigating external AI platforms. Victoria Usher, CEO of communications agency GingerMay, has gradually eased a blanket ban on ChatGPT, permitting its use with executive approval. Usher acknowledged that policies must remain flexible as AI becomes ingrained in work processes.
Yet, secrecy persists. A survey by Slack revealed that nearly half of desk workers hesitate to inform managers about their use of generative AI, fearing accusations of laziness or concerns over job security. Ethan Mollick, a management professor at the Wharton School, noted that employees often prefer to “look like geniuses” rather than disclose their reliance on AI.
The evolving legal landscape adds complexity to crafting long-term AI strategies. Uncertainties surrounding intellectual property, data privacy, and regulatory frameworks in regions like the U.S., EU, and UK have led some companies to ban AI use entirely. Michelle Roberts Gonzales, an employment lawyer at Hogan Lovells, told the Financial Times that such uncertainty can discourage innovation.
Rose Luckin, a professor at University College London, emphasized the need for companies to designate responsible teams for evaluating AI’s potential. In some cases, junior employees, like a paralegal at a London law firm, are unexpectedly tasked with designing AI rulebooks for senior staff.
