OpenAI reports that enterprise AI has officially moved out of the experimentation phase and into daily operational use. Businesses are no longer treating generative models as tools for basic tasks—they’re wiring them directly into core workflows, handing off increasingly complex sequences of work to the technology.
From Simple Prompts to Full Workflow Automation
While ChatGPT message volume has surged across the board, the more important indicator is the explosive growth in reasoning token consumption. OpenAI notes that API usage tied to logic-heavy tasks has grown nearly 320x per organisation, signaling that companies are now building AI into products and processes rather than using it solely for surface-level assistance.
Custom GPTs and Projects—tools that allow enterprises to encode internal knowledge—are also seeing significant adoption. Weekly usage has grown to nearly 19 times what it was last year, and about one-fifth of all enterprise messages now run through these customised systems. Standardisation is becoming essential to make AI usable at scale.
On the productivity side, OpenAI’s report points to consistent time savings across industries. Most workers claim the technology frees up 40–60 minutes each active day, while developers, analysts, and communications professionals report closer to 60–80 minutes.
Coding Skills Are Spreading Across Non-Technical Teams
One of the most notable shifts is the rise in coding-related tasks. OpenAI says that coding queries are increasing not just among engineers but across marketing, HR, operations, and finance teams. Over just six months, non-technical sectors recorded a 36 percent increase in coding-related usage, allowing these teams to solve problems that once required specialised developer support.
Operational metrics reflect similar improvements. Eighty-seven percent of IT professionals report faster issue resolution, and 75 percent of HR teams report improved employee engagement through AI-powered tools.
The Growing Divide Between Casual and Frontier adopters
OpenAI’s data shows a widening gap between organisations that simply give employees access to AI tools and those that build them directly into their operating models.
“Frontier” users—the top five percent in adoption intensity—send six times more messages than median users and engage in a broader range of tasks. Their organisations generate roughly twice the message volume per seat and seven times more interactions with custom GPTs.
The conclusion is clear: AI value correlates with depth of usage. Enterprises that keep AI deployment shallow see limited ROI, while those that integrate deeply and standardise workflows see exponential gains.
Industries are also adopting at different speeds. Tech remains the fastest-growing with an 11x jump year-over-year, but healthcare and manufacturing are accelerating quickly at 8x and 7x growth. Global adoption is also expanding beyond the US, with Australia, Brazil, France, and the Netherlands seeing more than 140 percent growth in business customers. Japan leads international API usage.
Real-World Examples: AI Driving Measurable Business Outcomes
Several deployments highlight how deeply integrated AI is reshaping operations:
• Retail: Lowe’s rolled out an AI-powered associate tool across more than 1,700 stores. When used, customer satisfaction increased by 200 basis points, and online conversion rates more than doubled.
• Pharmaceuticals: Moderna automated the creation of Target Product Profiles by letting AI extract insights from massive evidence packs. Tasks that once took weeks now take hours.
• Financial Services: BBVA built an AI system to handle routine legal queries around corporate signatory authority. It automatically resolves more than 9,000 queries per year—freeing the equivalent of three full-time employees.
Deep Integration Requires Organisational Readiness
The challenge for many enterprises is no longer the capability of the model—it’s the internal infrastructure and governance. About a quarter of companies still haven’t connected their internal systems or knowledge bases, limiting their models to general public information rather than company-specific context.
Successful organisations share similar traits:
• Executive support and clear mandates
• Secure access to internal data systems
• Standardised workflows and custom GPTs
• A culture that encourages using AI across tasks, not just for simple prompts
OpenAI’s findings suggest that the next wave of enterprise value will come from handing off complex, multi-step responsibilities—not just generating quick outputs. AI is evolving from a support tool into a core engine driving revenue, operational efficiency, and decision-making.
