Key Takeaways
- OpenAI’s models now serve more than 1 billion active users and over 2 million businesses, according to CFO Sarah Friar.
- Users deepen their AI engagement over time: individuals send 50 % more messages daily after six months and use ChatGPT for twice as many work types; businesses typically start with a single team before scaling enterprise‑wide.
- PYMNTS Intelligence shows generative AI becomes a mass habit through small, repeatable daily tasks such as finding product links, drafting texts/emails, and symptom lookup.
- In the enterprise AI race, financial services and insurance lead adoption, focusing on structured, auditable back‑office functions.
- OpenAI recently adjusted pricing and performance: GPT‑5.6 Luna price cut 80 %, GPT‑5.6 Terra price cut 20 %, and GPT‑5.6 Sol received faster API performance at unchanged cost, aiming to improve “intelligence, speed, reliability and cost” balance.
OpenAI’s Growing User Base
OpenAI’s technology has reached a milestone that underscores its rapid diffusion across the globe. Chief Financial Officer Sarah Friar revealed in a July 31 blog post that the company’s models now power more than 1 billion active users and are employed by over 2 million businesses. Friar emphasized that this scale is not merely a numeric achievement but a signal that AI is moving from experimental novelty to everyday utility. “We are still early,” she wrote, noting that the current adoption curve reflects only the beginning of what more capable systems will enable. The sheer breadth of users—from individual consumers to large enterprises—sets the stage for the next wave of AI‑driven productivity gains.
How Users Deepen Their AI Engagement
Friar’s post also highlighted behavioral shifts that occur as people grow more comfortable with the technology. Among individual users, after six months of consistent use, people tend to send 50 % more messages each day and expand the variety of tasks they delegate to ChatGPT, effectively doubling the kinds of work they rely on the model for. For businesses, the pattern is similar but operates at an organizational level: firms often begin by piloting AI with a single team or workflow, then gradually expand adoption across operations as confidence builds. This progression suggests that trust and familiarity are critical levers for unlocking deeper integration of AI into daily routines and corporate processes.
PYMNTS on the Habit‑Forming Power of Small Tasks
Supporting Friar’s observations, a PYMNTS Intelligence report titled “The AI On‑Ramp: Data Shows How Everyday Tasks Build Consumer Habits” argues that generative AI will become a mass habit not through grandiose projects but through small, repeatable actions performed daily. The report identifies three truly universal use cases: finding product links, drafting texts or emails, and symptom lookup. These activities are low‑friction, high‑frequency, and provide immediate utility, making them ideal entry points for habit formation. By repeatedly satisfying these minor needs, users develop a mental model of AI as a reliable assistant, paving the way for more complex applications over time.
Financial Services Lead Enterprise AI Adoption
Another PYMNTS study, “Financial Services Pull Ahead in the Enterprise AI Race,” reveals that while virtually every sector is experimenting with AI, financial services and insurance firms are among the earliest and most aggressive adopters. The report notes that these industries favor structured, auditable back‑office functions—such as transaction reconciliation, compliance monitoring, and risk modeling—where AI’s ability to process large volumes of data with precision delivers clear ROI. The emphasis on auditability reflects the sector’s regulatory constraints, suggesting that AI tools that can provide transparent, traceable outputs are particularly valued in this space.
Recent Model Pricing and Performance Adjustments
Friar’s blog appeared just one day after OpenAI announced a series of price and performance tweaks aimed at improving the cost‑efficiency of its API offerings for enterprise workloads. The company cut the price of GPT‑5.6 Luna by 80 %, reduced the cost of GPT‑5.6 Terra by 20 %, and enhanced the speed of GPT‑5.6 Sol while leaving its price unchanged. These adjustments were framed not as arbitrary discounts but as strategic moves to expand the range of work that becomes practical for customers. Friar explained, “These are not simply changes to a price list. They expand the range of work that becomes practical and give customers more flexibility to balance intelligence, speed, reliability and cost.”
Strategic Rationale Behind the Changes
The pricing revisions signal OpenAI’s intent to lower barriers for broader experimentation, especially among cost‑sensitive startups and mid‑market firms that may have previously hesitated to invest heavily in AI inference. By making Luna dramatically cheaper, OpenAI enables high‑volume, latency‑tolerant tasks—such as batch content generation or large‑scale data enrichment—to be run at a fraction of prior expense. The modest Terra reduction targets workloads that need a balance of capability and affordability, while the Sol performance boost addresses latency‑critical applications like real‑time chatbots or interactive coding assistants without raising costs. Collectively, these changes aim to shift the economics of AI usage so that organizations can allocate more budget toward innovation rather than infrastructure.
Implications for Individual Users and Small Businesses
For individual consumers, the price cuts translate into greater access to powerful models for everyday tasks like drafting emails, brainstorming ideas, or seeking quick information—activities highlighted in the PYMNTS habit‑forming report. Small businesses, which often operate on tight margins, can now afford to embed AI into multiple functions simultaneously, from customer support automation to marketing copy generation. Friar’s assertion that “individuals and small businesses will gain capabilities once available only to much larger organizations” becomes more plausible” reflects this democratizing effect. As the cost of using state‑of‑the‑art models declines, the gap between enterprise‑grade AI tools and those accessible to the everyday user narrows.
Enterprise‑Scale Impact and Future Outlook
Large enterprises stand to benefit from the performance improvements to Sol, which allow them to run more sophisticated, low‑latency AI services at existing cost levels. This is particularly relevant for industries such as finance and healthcare, where real‑time decision‑making is crucial. Friar’s vision of “more capable systems [that] will complete longer projects, coordinate across tools, and handle more of the work between an idea and a finished result” aligns with the trend toward AI agents that can orchestrate multiple APIs, manage workflows, and produce end‑to‑end outputs with minimal human intervention. The current pricing and performance adjustments are therefore early steps toward enabling such orchestration at scale.
Conclusion: A Maturing AI Ecosystem
OpenAI’s latest disclosures paint a picture of a maturing AI ecosystem where scale, usage depth, and economic accessibility reinforce one another. With over a billion users and two million businesses already on board, the company is witnessing the transition from sporadic experimentation to entrenched, habit‑driven utilization. The PYMNTS insights reinforce that this entrenchment begins with modest, daily tasks, while enterprise adoption—especially in finance—shows where AI delivers the highest immediate value. By recalibrating price and performance, OpenAI is seeking to widen the aperture for both exploratory use and mission‑critical deployment, setting the stage for the next phase of AI‑enabled productivity across the globe.
OpenAI Reaches 1 Billion Active Users as AI Becomes Daily Habit