Need to Know

• There is broad agreement that AI will have a major impact, but less certainty about what that impact will look like.
• Companies are investing in AI even as they work to demonstrate its financial value.
• Practical applications are emerging across chemical manufacturing and molecular research.
• For the adhesives and sealants industry, the conversation should shift from whether AI matters to where it can be useful.

Companies know they need AI, but most aren't seeing significant financial benefits yet. That may be because AI requires a longer investment horizon and deeper organizational changes than previous technologies. Executives themselves say the biggest barrier isn't technology. It is people and organizational change.

In January, PwC released a study reporting that 42% of the CEOs surveyed listed concerns about whether they are “transforming fast enough to keep pace with technological change” as their top concern. However, only 12% reported that AI had resulted in both “cost and revenue benefits,” while 56% reported “no significant benefit.” Mohamed Kande, PwC global chairman, saw the results as an indication of a widening gap between companies able to translate AI use into financial results and those that have not.

At the end of 2025, Deloitte released a survey of 1,854 executives across Europe and the Middle East describing “the paradox of rising investment and elusive returns.” While participants viewed AI as a strategic necessity, companies reported a return on investment for AI use cases within two to four years, much longer than the seven- to 12-month payback period of other technology investments. Despite that, executives believe in AI’s long-term potential and remain committed to its implementation. Deloitte compared the challenges of adopting AI to those manufacturers faced when switching factories from steam power to electricity, forcing them to “reconfigure their production lines, redesign workflows, invest in new infrastructure, and reskill their workforce.”

In its 2026 AI & Data Leadership Executive Benchmark Survey, Harvard reported that business leaders are bullish about the technology. The survey of nearly 110 companies, with many participants working as data and AI executives, found that investment in data and AI is “nearly universal.” In fact, 99.1% reported investment in data and AI as a “top organizational priority.” More interesting to me, 93.2% reported that the greatest impediment to AI adoption was not technology but “people and change,” as organizations seek “cultural transformation,” new business processes and education. This tracks with my own experiences listening to people in my orbit discuss AI and its usefulness.

What does this mean to the adhesives and sealants industry? To begin with, industry needs to move the conversation from whether AI matters to where it can be useful. One way companies can overcome the natural human resistance to change is to show people how that change can create progress and success. Leadership also needs to encourage creativity along with practical thinking when considering use cases for AI technology.

The chemical industry is already providing examples of how AI can change established ways of working, much as electric power eventually changed the operation of steam-powered factories. Celanese offers one example. The company introduced JO.AI, a platform with specialized, task-specific AI agents that perform tasks such as optimizing shift checklists, flagging equipment anomalies, and drafting work orders. The tool integrates industrial data from more than 40 sources, including sensors, applications and images, and uses that information to map relationships between equipment and people. In a different use case, Celanese launched AskChemille.com last year. Chemille is a customer-facing digital assistant that helps design engineers access material-selection data, providing predictive insights and customized recommendations based on customers’ specifications and requirements.

Another tool available to the chemical industry is PNNL’s Chemistry Agent Connecting Tool-Usage to Science (CACTUS). Designed to assist researchers and chemists with molecular discovery involving therapeutic candidates, catalysts and materials, the AI-powered tool connects large language models with computational chemistry and cheminformatics tools. PNNL maintains a public GitHub repository containing the CACTUS source code, documentation, tutorials, tests, and benchmarking resources. The underlying CACTUS research is peer-reviewed and was published in the November 2024 issue of ACS Omega.

As someone close to many people with advanced degrees, Wolfers’ discussion of AI’s impact on those who do “cognitive work” causes some concern, as I am sure it does for scientists working in our industry. After looking into how chemical companies and research institutes are using AI, however, I think there is an argument to be made that the more immediate challenge for our industry is figuring out how expertise and AI can work together productively.

Thanks for reading this issue of ASI, and as always, please contact me at parkerk@bnpmedia.com with comments and questions.