Do you remember the first time you used ChatGPT? It probably wasn’t Nov. 30, 2022, when the platform debuted. But it was probably soon after. More than 800 million weekly users later it’s on track to have more than 220 million paid users by 2030.
ChatGPT and the other large language models (LLMs) that came after in short order have changed pretty much everything in the business world and as PYMNTS Intelligence research showed, it was a factor on Black Friday. Half of all shoppers, including two in three Gen Zers, used conversational artificial intelligence (AI) tools to help them complete their purchases.
But the big story three years into its life is scale. PYMNTS data showed that 90% of CFOs reported very positive ROI from adopting generative AI, up sharply from 27% in March 2024, reflecting a decisive shift from pilot experiments to structured deployment. OpenAI added one million paying business subscribers between February and June 2025, bringing the total to three million across ChatGPT Enterprise, Team and Education tiers, according to PYMNTS. ChatGPT Plus maintains the highest retention rate in the category, with 71% of users continuing after six months.
And for the financial services industry, which has been inconsistent in its digital transformation, generative AI is a genuine break with the past. PYMNTS CEO Karen Webster called it “the technology that broke the adoption curve” because it bypassed the usual frictions. It needed no new hardware, terminal upgrades or network rollouts. People could simply start typing. That ease of use is resetting expectations for what digital should feel like in banking, payments, fintech and retail.
Webster argued that, when tied to the right problems, the value is more than incremental. PYMNTS Intelligence research, she noted, showed “a strong — or what I might even call compelling — positive impact from generative AI in the corporate world.”
From Pilots to Deployment
A PYMNTS study revealed that companies are now treating AI as an essential part of their workforce, rather than just an experiment. Sixty large U.S. firms are reorganizing roles and responsibilities under new Chief AI Officers.
The data showed that 34% of CFOs cite increased output as their top reason for adopting AI according to PYMNTS. Staying competitive comes next, at 24%, followed by improved decision-making via better data insights, at 19%.
The data also revealed sharp industry divides. 48% of goods sector firms are using AI to boost output and efficiency. In contrast, 30% of service firms aimed to improve decision-making and customer experience with AI. Additionally, 42% of technology firms stated that their main goal is to maintain their competitive edge.
Workforce Transformation
60% of CFOs in a PYMNTS study said their firms are at least somewhat ready to manage AI-driven change, including just 12% who believe they are very prepared. Nearly all of the rest (38%) remained neutral, indicating a lack of confidence about their ability to adapt.
Executives from various sectors identified human factors as the main challenges. They pointed to skill gaps, cultural resistance, and difficulties in training.
AI adoption readiness levels vary widely. 75% of tech companies felt at least somewhat prepared to handle AI’s effects on the workforce. This compared to 63% of goods producers and only 48% of service firms. Companies also follow different operational strategies. In the goods sector, 30% of CFOs prioritized hiring talent with AI skills, while 26% focused on reducing staff as automation increases. Service firms tended to favor selective task automation at 25% and redesigning roles at 20%. Technology companies used a combination of upskilling, outsourcing, and targeted hiring as needed.
Measuring Returns Beyond Cost Reduction
Enterprises now monitor generative AI with tighter financial discipline. CFOs assess GenAI based on revenue gains, cost savings, cost avoidance, and risk reduction. They no longer view AI as a side tool. Instead, they incorporate it into long-term performance planning and use it to improve decision-making, operational resilience, and competitive positioning.
PYMNTS data showed clear differences in ROI across industries. Information-sector firms lead the market, with 65% reporting very positive ROI because they invest heavily in customized models and advanced automation systems.
Manufacturing firms follow. 56% reported active use of generative AI for production diagnostics, predictive maintenance and factory-floor optimization. These firms use AI to reduce quality issues, speed up anomaly detection and stabilize output. Roughly 1 in 3 construction companies use the technology to spot fraud or flag inconsistencies. Further analysis showed that 44% of construction firms reported the technology helps with product or service innovation.
The retail sector was quick to deploy generative AI, with 71% of companies implementing it for customer-facing chatbots. These bots are crucial for handling a large number of interactions. Customer engagement affects business results in retail, so chatbots are significant tools. As they improve service delivery, personalize customer interactions, and boost operational efficiency.
But retailers showed the widest performance gap. They rely heavily on conversational agents for customer engagement, but only 17% reported very positive ROI. The shortfall might be due to retailers’ dependence on baseline models and limited integration into core merchandising and inventory workflows.
Adoption is expanding across financial functions as well. PYMNTS Intelligence found that 82% of CFOs use or are exploring AI in accounts payable. Similar adoption patterns appear in cybersecurity, procurement, financial modeling and customer operations as generative AI becomes embedded in enterprise digital operating systems.
Still, it is not all green. Enterprises face real risks, from uneven workforce readiness to gaps in model reliability and security. Some industries continue to struggle with low ROI, and many firms remain unsure about how quickly they can adapt. But these challenges are becoming more manageable as organizations gain experience and build the structures needed to deploy AI with greater control and discipline.
Source: https://www.pymnts.com/