This month in AI: ChatGPT usage patterns, governance gaps and mega-infrastructure bets

New data from OpenAI shows how the use of ChatGPT has moved from experimentation to the core of daily work and life.

In our monthly roundup of AI insights, we unpack what these latest generative AI usage patterns tell us about work, highlight practical actions from the Forum’s new playbook on responsible AI and track the latest governance and infrastructure moves shaping the ecosystem.

Plus, a chart on surging global AI spending and curated must-reads to help you cut through the noise.

OpenAI, the maker of ChatGPT, recently published findings from what it calls 'the largest study to date' on how people are using the chatbot. The results show a subtle but important shift: work-related activity now makes up a smaller share of ChatGPT use than it did a year ago.

That’s not because fewer people are using the tool at work — the study suggests both work and non-work messages have grown continuously. Instead, non-work uses are growing faster, now representing more than 70% of all consumer messages.

When it comes to work use, OpenAI finds that 81% of messages are centred around seven activities, led by documenting information (18.4%) and making decisions or solving problems (14.9%).

These patterns hold across occupations, from management and STEM to sales and administration — suggesting generative AI is transforming tasks more than entire jobs.

Meanwhile, researchers are cautioning against the rise of "workslop", or low-effort AI-generated output that appears polished but lacks real substance. Coined by BetterUp Labs in collaboration with Stanford's Social Media Lab, the term describes content that shifts the burden downstream: where colleagues must interpret, correct or redo the work.

Their survey of 1,150 US employees found that 40% had received "workslop" in the past month, with each incident costing nearly two hours of lost time. It's being seen as an invisible "tax" worth an estimated $9 million annually for a 10,000-person organization. Beyond productivity, the study warns that "workslop" erodes trust and collaboration among teams.

Nearly three years into the "generative AI moment," this shift is exactly what the Forum’s new report, Advancing Responsible AI Innovation: A Playbook, responds to.

While consumer and workplace adoption accelerates, the Playbook sets out practical steps for business and government leaders to ensure AI's impact is safe, responsible and widely beneficial.

The report notes that maturity in implementation lags behind awareness of its importance, creating a "responsible AI implementation gap." If left unaddressed, "this lapse in AI governance is likely to erode confident AI investment, compliance and public trust."

Dig Deeper: