Chemical producers with an AI-first mindset are already delivering significant value. From finding high-potential molecules at a specialty producer to maximizing operational margins, AI is improving profitability and enhancing competitive edge.
AI’s impact is much needed in an industry facing a lengthy list of challenges, including structural overcapacity, weak demand in key markets, and intense pressure from customers and regulators to decarbonize. Geopolitical instability adds extra uncertainty and complicates strategic planning.
Nevertheless, CEOs at some producers are driving enterprise-wide AI investment. They see AI as a critical tool for managing today’s cost pressures—and as a source of strategic advantage.
Commodities producers driving a broad AI transformation can see an EBITDA uplift of 3% to 5%; specialty players even more, at 4% to 6.5%. More strategically, when the cyclical upturn comes, AI-first producers will have the agility, data, and operating models to capture a larger share of the upswing.
How AI Has Already Generated Value
A specialty producer unlocked an R&D breakthrough. This Europe-based producer needed to identify a solvent with a highly specific, hard-to-optimize set of properties for a carbon capture process. BCG helped the company use AI to evaluate 100 times more candidate molecules than human researchers could realistically screen, quickly identifying the most promising candidates. Compared with initial expectations, the solvent enabled a more efficient process and was less corrosive, enabling the use of cheaper construction materials.
The result: opex savings of 5% to 10% and a capex reduction of 10%.
A petrochemical company boosted gross margins through plant optimization. This Middle Eastern producer was struggling to manage hundreds of process variables. An AI/machine learning model established the parameters that yielded the best economic results, and an AI agent implemented them automatically, with operator override. Gross margins were boosted by $8 to $15 per tonne.
A large chemical distributor boosted margins and found new growth. This global distributor faced structural issues in its commercial/sales model, including 15,000 products, a fragmented customer base, and inefficient internal handoffs. BCG helped the company build a series of AI models that identify cross-selling opportunities and provide sales reps with a single-screen view, along with real-time coaching and next-best actions. Each rep is now generating 30 to 40 more leads a month, and EBITDA is up 3% to 5%.
Stay ahead with BCG insights on industrial goods
The Key: Deploy, Reshape, and Invent
Without strong CEO oversight, many chemical companies remain stuck at the first stage of AI implementation—capturing only a fraction of AI’s potential value.
Deploy. At this stage, AI tools are used to boost productivity in isolated tasks, such as automated testing on a batch production line.
Reshape. With strategic guidance from the top, companies at this stage are reimagining end-to-end processes and workflows to be AI-enabled. Not just automated testing, for instance, but a full diagnostic system that identifies problems, suggests likely causes, and optimizes to drive down failure rates.
Invent. This involves building AI-native offerings that enhance customer value or create new business models. For instance, a company that produces a chemical treatment could develop AI tools that tell customers the optimal time and quantity to apply it. Fees could be linked to customer outcomes such as improved quality or reduced downtime.
The Strategy Difference
The difference between AI leaders and laggards is not primarily technical—it is strategic, driven by the CEO.
For commodity volume players, about half of their potential 3% to 5% EBITDA uplift comes from operations. But CEOs who drive a wider deployment can also achieve valuable improvement in supply chain optimization and other commercial functions, potentially doubling gains.
For specialty companies, the benefits of a broad, CEO-led AI transformation are even greater. AI in operations can give an EBITDA uplift of 1.2% to 1.8%—valuable, but only a part of the improvement. There are also valuable gains from optimizing commercial operations, such as pricing power and product mix, and from improving supply chain management.
The total expected 4% to 6.5% EBITDA improvement from end-to-end transformation can be further boosted by the substantial, case-by-case impact of using AI to discover new molecules and products.
Becoming an AI-First Chemical Company
To become AI-first, companies must:
- Pursue a phased, multi-year ambition. An active, engaged C-suite is essential for this, often backed by the appointment of chief AI and data officers to drive implementation.
- Move to an AI-first operating model. Drive partnerships between IT and the business to implement AI. Ensure that guardrails and governance are strong.
- Focus on reshape and invent. The yardstick should be P&L impact, not the number of tools in operation. Strong value tracking provides an essential financial framework.
- Use fit-for-purpose technology and data. An enterprise AI platform drives efficient rollout. Templates and enterprise-wide data models also speed deployment at scale.
- Find the necessary talent—and enable it. Structured AI learning programs are part of this. But it’s also vital to involve employees in shaping and adopting AI workflows. In the case of the distributor described earlier, effective culture management meant 98% of sales reps were satisfied with the system, ensuring key sales talent was retained.
With budgets under pressure and many producers entering another round of cost-cutting—especially in petrochemicals—some CEOs may feel they cannot afford enterprise-wide AI investment. In reality, they cannot afford to delay it: AI-first companies are already positioning themselves to win the next cycle before demand recovers.
