Dow

Experimenting
  • Chemicals
  • 31,600 staff (Sep 2026)
  • Copilot, C3 AI, OpenAI chat
  • 4,500 roles cut under AI-linked plan
  • No AI return reported

Dow deploys AI through bought tools on its own data: Microsoft 365 Copilot and Copilot Studio agents, C3 AI predictive maintenance on more than 50 steam cracker furnace operations, and an OpenAI-assisted market intelligence hub, all fed by an Integrated Data Hub on Azure and Databricks. Dow tied its Jan 2026 Transform to Outperform plan, about 4,500 role cuts and at least $2 billion of EBITDA uplift, partly to AI and automation, but has not said how much of the roughly $450 million delivered by Sep 2026 came from AI.

corporate.dow.com

Their approach to AI

The platform
Dow buys its AI tools and runs them on its own Integrated Data Hub, built on Microsoft Azure and Databricks, which Chief Data & Analytics Officer Chris Bruman called the biggest change to Dow's AI strategy (Apr 2025).
The models
Dow uses Microsoft 365 Copilot and agents built in Copilot Studio, C3 AI Reliability for equipment, and OpenAI-assisted chat in its Market Intelligence Hub; Bruman said staff may use only pre-approved large language models (Apr 2025).
The reach
Microsoft reported in Mar 2025 that Dow had extended Copilot to nearly 60% of the company, up from about 20% in Nov 2024; Dow itself has not published a user count.
The catch
Dow credits its Transform to Outperform savings to AI and automation together with process redesign, and CFO Jeff Tate said in Sep 2026 that most of the work is process optimization and automation, with AI in some workflows; Dow has published no figure for AI alone.

Snapshot

Employees
~31,600
Sep 30, 2026; about 34,600 at Dec 31, 2025
Technology budget
Not disclosed
R&D expense was $752M in 2025, down from $810M in 2024
AI/ML headcount
Not disclosed
Enterprise Data & Analytics runs a Delivery Center for AI and analytics (Apr 2026)
Who owns AI
Debra Bauler, CIDO
Information technology sits under CFO Jeff Tate (2026 proxy)
Platform
Integrated Data Hub
Built on Microsoft Azure and Databricks; tools are bought
Models
Copilot, OpenAI chat, C3 AI
Only pre-approved large language models allowed (Apr 2025)
Reach
Nearly 60% on Copilotest.
Microsoft, Mar 2025; no Dow-reported count
AI-linked program
Transform to Outperform
About 4,500 roles and ~$2B of EBITDA uplift, announced Jan 2026

Key findings

  1. Dow announced Transform to Outperform on Jan 29, 2026, a plan to cut about 4,500 roles and add at least $2 billion of Operating EBITDA, in part by using AI and automation, and delivered about $450 million of that in the first 9 months of 2026 without saying how much came from AI.

  2. Jim Fitterling told analysts on Jan 29, 2026 that Transform to Outperform is not all AI, and CFO Jeff Tate said in Sep 2026 that Dow relies mainly on process optimization, standardization and automation, with AI helping some workflows.

  3. Dow runs C3 AI Reliability on more than 50 steam cracker furnace operations and in Jun 2025 let its subsidiary Univation Technologies license the same software to other petrochemical producers.

  4. Microsoft reported in Mar 2025 that Dow had extended Microsoft 365 Copilot to nearly 60% of the company and built 2 freight invoice agents in Copilot Studio, from which Dow anticipated millions of dollars in savings; Dow has not reported the realized figure.

  5. Dow opened a Delivery Center for AI and analytics at the University of Illinois Research Park on Apr 8, 2026, inside its Enterprise Data & Analytics organization, with AI governance as one of its 4 pillars.

Where Dow sits

Stage
Experimenting

Dow has reported no business return it credits to AI since May 2025. The roughly $450 million of Operating EBITDA improvement Dow delivered from Transform to Outperform in the first 9 months of 2026 (Q3 2026 Form 10-Q) comes from a program the company says works in part through AI and automation, alongside role cuts, site playbooks, sourcing and pricing; Jim Fitterling said on Jan 29, 2026 that it is not all AI, and CFO Jeff Tate said in Sep 2026 that most of it is process optimization and automation. Fitterling's remark that AI brings big cost savings in legal patent and discovery work, and Kyndryl's Nov 2025 statement that their collaboration delivered measurable cost savings, carry no figure. On rollout, Dow has not itself reported how many employees use its AI tools: the nearly 60% Copilot figure comes from a Microsoft blog in Mar 2025, before the window. Dow's AI work is on the recent record, from the C3 AI expansion (Jun 2025) to the Delivery Center for AI and analytics (Apr 2026), so Dow sits at Experimenting rather than Unknown. A return Dow credits to AI in one use case, such as predictive maintenance or freight invoice auditing, or a company-reported AI user count under Chief Information & Digital Officer Debra Bauler, would move Dow to Piloting.

How the stages are defined
Ownership
Chief Information & Digital Officer Debra Bauler, under CFO Jeff Tate
Platform
Bought: Bought tools on an in-house data hub (Azure, Databricks)
Models
Microsoft Copilot, OpenAI-assisted chat, C3 AI Reliability
Hosting
Integrated Data Hub on Microsoft Azure and Databricks; Copilot in Microsoft's cloud; Kyndryl runs Dow's IT infrastructure; where the OpenAI chat and C3 AI run is not disclosed
Reach
Nearly 60% on Copilot (Microsoft, Mar 2025)

How they deploy

Data first, then AI

Dow built its Integrated Data Hub on Microsoft Azure and Databricks to put data from almost every function and business in one place, with rules on data product ownership, metadata and quality, and won a 2024 CIO 100 award for it. Chris Bruman, Dow's Chief Data & Analytics Officer, said in Apr 2025 that the hub was the most significant change to Dow's AI strategy, and Jim Fitterling said on Jan 29, 2026 that AI on top of the hub will let Dow take steps out of its work processes.

A few generative AI bets

Chris Bruman said in Apr 2025 that Dow had avoided launching dozens of generative AI projects and focused on a handful of high-impact problems, such as parsing long documents, and that patent reviews for new research had fallen from months to weeks or days. Dow has used AI in chemistry and engineering for more than 2 decades, Bruman said, and its predictive models cover demand forecasting, pricing, delivery issues and equipment health.

AI inside a cost program

Dow framed Transform to Outperform in Jan 2026 as a redesign of its operating model and end-to-end processes, with about 2 thirds of the benefit from productivity and one third from growth, partly through AI and automation. Jim Fitterling said the program is not all AI and named procurement, third-party costs and process handoffs as targets, and CFO Jeff Tate said in Sep 2026 that Dow had found more than 50 opportunities to redesign end-to-end processes and was focused on eliminating and automating work.

Buy the tools, keep the data

Dow buys its AI software from Microsoft, C3 AI and others instead of building models, and Chris Bruman said in Apr 2025 that Dow keeps critical data within the company, lets employees use only pre-approved large language models with data protection, and counts on its strategic partners to indemnify it against copyright claims.

Deployment model
Dow runs AI through a central Information Technology organization led by Chief Information & Digital Officer Debra Bauler, who oversees information systems, AI, data analytics, intelligent automation and cybersecurity. Chief Data & Analytics Officer Chris Bruman leads the centralized Enterprise Data & Analytics team, chairs a Data & Analytics Council of senior business and functional leaders, and is moving reporting to a hub-and-spoke model in which business users build their own analytics on the Integrated Data Hub. Transform to Outperform runs through a Transformation Management Office led by Mike Weideman, with end-to-end process owners for workflows such as plan-to-fulfill (Apr 2026).
Model providers
Dow uses Microsoft 365 Copilot, which began with a few hundred employees and reached about 20% of the workforce by Nov 2024 and nearly 60% of the company by Mar 2025, according to Microsoft, plus agents built in Microsoft Copilot Studio. Dow's Market Intelligence Hub includes OpenAI-assisted chat and AI summaries of news and reports (Mar 2025). For equipment, Dow runs C3 AI Reliability on more than 50 steam cracker furnace operations (Jun 2025). Chris Bruman said in Apr 2025 that employees may use only pre-approved large language models; Dow has not named which models those are beyond these tools.
Data and infrastructure
Dow's Integrated Data Hub runs on Microsoft Azure and Databricks and holds data from almost every function and business (Apr 2025); the Freight Agent was trained on 43,000 invoices from it (Mar 2025), and Dow's Carbon Footprint Ledger, built on Databricks with Boston Consulting Group, sits on it too (Apr 2026). Kyndryl has worked with Dow for nearly 2 decades and manages its IT infrastructure, including cloud, network, digital workplace and security services, and agreed in Nov 2025 to modernize Dow's application landscape with AI and automation. Dow has not said where its OpenAI-assisted chat or its C3 AI deployment is hosted.

Platforms and internal tools

  • Integrated Data HubBy Mar 2024

    The Integrated Data Hub is Dow's enterprise data platform on Microsoft Azure and Databricks, with domain-oriented data landing zones, automated metadata and a data marketplace through which data owners control access. Dow paired it with a company-wide data and analytics literacy program, and Dow's AI tools, the Freight Agent and the Carbon Footprint Ledger draw on it.

    Source

  • Microsoft 365 Copilot and Copilot Studio agents2024

    Dow rolled out Microsoft 365 Copilot from a few hundred employees to about 20% of its workforce by Nov 2024, according to Microsoft, which reported nearly 60% in Mar 2025. In logistics, Dow built an autonomous agent that reads PDF freight invoices from email and checks them against negotiated rates, and a Freight Agent for natural-language questions, at the proof-of-concept pilot stage in Mar 2025, starting with North American road freight. Microsoft said Dow was also building an agent to help R&D with patenting.

    Users
    About 20% of the workforce (Nov 2024); nearly 60% of the company (Mar 2025), per Microsoft
    Models
    Microsoft 365 Copilot

    Source

  • Market Intelligence HubBy Mar 2025

    Dow's Market Intelligence Hub gives employees market data, customizable analysis, AI-written summaries of news and reports, targeted email subscriptions, and an OpenAI-assisted chatbot over business-curated intelligence and external web data. Dow won a 2025 CIO 100 award for it, its fifth consecutive CIO 100 win.

    Models
    OpenAI (assisted chat)

    Source

  • C3 AI Reliability

    Dow uses C3 AI Reliability, part of the C3 AI Asset Performance Suite, for predictive maintenance on more than 50 steam cracker furnace operations, and is extending it to more furnaces and other critical assets. Debra Bauler said in Jun 2025 that C3 AI had moved Dow's operations from reactive to predictive and that Dow anticipates significant benefits as it scales; Dow's subsidiary Univation Technologies now licenses the software to other producers.

    Source

  • Dow Delivery CenterApr 8, 2026

    Dow's Delivery Center at the University of Illinois Research Park in Champaign is a hub for AI and analytics inside the Enterprise Data & Analytics organization, built around data excellence, business strategy, applied use cases and AI governance, and staffed by full-time professionals and University of Illinois students. Jen Quinlan leads the site.

    Source

Vendors and partners

  • MicrosoftCopilot by 2024

    Azure cloud under the Integrated Data Hub; Microsoft 365 Copilot and Copilot Studio · Microsoft's Nov 2024 case study and Mar 2025 blog are the source of Dow's Copilot reach figures and of the freight agent work.

  • Databricks

    Data platform under the Integrated Data Hub and the Carbon Footprint Ledger

  • C3 AI

    C3 AI Reliability for predictive maintenance on steam cracker furnaces · Univation Technologies, a Dow subsidiary, agreed on Jun 30, 2025 to license and distribute the C3 AI software Dow uses.

  • OpenAIBy Mar 2025

    Model behind the Market Intelligence Hub's assisted chat

  • KyndrylNearly 2 decades; expanded Nov 10, 2025

    Manages Dow's IT infrastructure; modernizing Dow's applications with AI and automation

  • Boston Consulting Group

    Built the Carbon Footprint Ledger with Dow on Databricks

Leadership

AI, data and platform leaders

DB

Debra Bauler

Chief Information & Digital Officer

Debra Bauler joined Dow in 2023 as vice president of Global IT after more than 28 years at Cargill, where she was a vice president of IT and Global Group CIO, and held the chief information and digital officer role by Mar 2025. Bauler sits on Dow's Leadership Team and leads digital transformation and data strategy across information systems, AI, data analytics, intelligent automation and cybersecurity.

CB

Chris Bruman

Chief Data & Analytics Officer

Chris Bruman, a Dow employee of more than 30 years, leads the Enterprise Data & Analytics organization, chairs Dow's Data & Analytics Council and is accountable for AI, machine learning and self-service analytics at scale; he is also business unit CIO for Industrial Intermediates & Infrastructure. Bruman was quoted as Chief Data Analytics Officer at the Delivery Center opening on Apr 8, 2026.

MW

Mike Weideman

Head of Enterprise Transformation

Mike Weideman leads Dow's Transformation Management Office and Transform to Outperform. Weideman joined Dow's corporate IT in 1995, later ran global corporate and professional services procurement and an HR employee services team, and most recently led Enterprise Architecture and digital innovation in IT, where in Nov 2024 he was quoted as a senior IT director on Copilot.

Executives

KC

Karen S. Carter

Chief Executive Officer

Karen S. Carter became CEO of Dow on Jul 1, 2026, after serving as Chief Operating Officer from Dec 2024, President of Packaging & Specialty Plastics, and Chief Human Resources Officer. Carter joined Dow in 1994, announced Transform to Outperform as COO in Jan 2026, and led its activation, including setting up the Transformation Management Office, according to the 2026 proxy.

JF

Jim Fitterling

Executive Chair

Jim Fitterling was CEO of Dow from 2018 to Jul 1, 2026 and chair from Apr 2020, and is now Executive Chair. Fitterling said on the Jan 29, 2026 earnings call that rapid advances in AI and automation were one reason for Transform to Outperform, that the program is not all AI, and that AI is already cutting costs in legal patent research and case discovery.

JT

Jeff Tate

Chief Financial Officer

Jeff Tate rejoined Dow as CFO in Nov 2023 from Leggett & Platt, where he was CFO, after earlier Dow roles including chief audit executive. Tate has executive oversight for finance, integrated supply chain, purchasing and information technology (2026 proxy), and said in Sep 2026 that Transform to Outperform relies mainly on process optimization and automation, with AI in some workflows.

AA

Andre Argenton

Chief Technology & Sustainability Officer

Andre Argenton became Chief Technology & Sustainability Officer in Jan 2026, leading Dow's global R&D organization as well as sustainability and environment, health and safety. Argenton joined Dow in 1999 and was Chief Sustainability Officer from 2022 and vice president of R&D from 2018 to 2022.

AI org map

Central IT and data office running bought AI tools; transformation office driving the AI-linked cost program

Dow's Information Technology organization is led by Chief Information & Digital Officer Debra Bauler, a member of the Leadership Team. The 2025 Form 10-K says the CIDO reports to the chief operating officer, while the 2026 proxy and Jeff Tate's Dow bio give CFO Tate executive oversight of information technology. Chris Bruman's Enterprise Data & Analytics organization holds the Integrated Data Hub and the Delivery Center, and Mike Weideman's Transformation Management Office runs Transform to Outperform, which Karen S. Carter set up as COO and now oversees as CEO.

Dow's AI work sits in the Information Technology organization under Chief Information & Digital Officer Debra Bauler, in Chris Bruman's centralized Enterprise Data & Analytics team, which now includes the Delivery Center for AI and analytics in Champaign, Illinois (Apr 2026), and among R&D scientists who publish on machine learning for chemistry. Dow trains employees through a Citizen Data Science program, described in Digital Discovery in 2025, and a data and analytics literacy program. Dow has not published how many data scientists or AI engineers it employs.

Reporting lines

  • Chief Executive Officer

    Karen S. Carter

    • Finance and Information Technology

      Jeff Tate, CFO

      Also integrated supply chain, purchasing and investor relations (2026 proxy)

      • Information Technology and digital

        Debra Bauler, CIDO

        The 2025 Form 10-K gave the COO as the CIDO's manager; the 2026 proxy places IT under the CFO

        • Enterprise Data & Analytics

          Chris Bruman, CDAO

          Reporting line inferred from the CIDO's remit; not published

    • Transformation Management Office

      Mike Weideman

      Reporting line not published; Karen S. Carter set the office up as COO

    • Research & Development, Sustainability and EH&S

      Andre Argenton, CTSO

Dow has not published an org chart for IT or data. The executive officers listed in the 2025 Form 10-K (Feb 2026) do not include the CIDO, while Dow's leadership page lists Debra Bauler and Mike Weideman on the Leadership Team (checked Nov 5, 2026).

Where the AI workforce sits

  • Enterprise Data & Analytics (central)

    Dow has published no count of its data and AI staff; the Delivery Center adds students and full-time staff in Champaign.

    Not disclosed
  • R&D and business-unit data science

    Dow R&D scientists publish machine learning work and Dow runs a Citizen Data Science program for staff outside the central team.

    Not disclosed

Named people

Executive sponsors
  • Karen S. Carter

    Chief Executive Officer

  • Jim Fitterling

    Executive Chair

  • Jeff Tate

    Chief Financial Officer · Executive oversight of information technology (2026 proxy)

Information Technology and digital
  • Debra Bauler

    Chief Information & Digital Officer

Enterprise Data & Analytics

Runs the Integrated Data Hub, the Data & Analytics Council and the Delivery Center.

Transformation Management Office
Research & Development
  • Andre Argenton

    Chief Technology & Sustainability Officer

Use cases

Dow has described AI use cases in production or pilot from C3 AI predictive maintenance on more than 50 steam cracker furnace operations and Microsoft Copilot agents that audit freight invoices to OpenAI-assisted market intelligence, generative AI for patent review and legal discovery, and predictive models for demand, pricing and delivery. Dow has not published a business return it credits to AI for any of them. At firm level, Dow delivered about $450 million of Operating EBITDA improvement from Transform to Outperform in the first 9 months of 2026 and targets about $2 billion by the end of 2027, but credits that to AI and automation together with process redesign, role cuts and sourcing, with no AI share.

Use caseAreaStatusROI
Predictive maintenance on steam cracker furnaces

Dow uses C3 AI Reliability to predict equipment problems on more than 50 steam cracker furnace operations and is extending it to more furnaces and other critical asset classes; Debra Bauler said in Jun 2025 that Dow anticipates significant benefits as it scales.

ManufacturingIn production
Drones and crawlers in plant turnarounds

Jim Fitterling said on Jan 29, 2026 that drones and crawlers with cameras, part of what he called traditional AI, let Dow avoid much of the scaffolding cost of plant turnarounds and make them safer.

ManufacturingIn production
Freight invoice auditing agents

Dow built an autonomous Copilot Studio agent that reads PDF freight invoices from email and flags mismatches with negotiated rates, and a Freight Agent trained on 43,000 invoices for questions in natural language, starting with North American road freight; Microsoft described the work as a proof-of-concept pilot in Mar 2025, with Dow anticipating millions of dollars of savings once scaled.

Supply chainPilot
Demand forecasting

Dow uses predictive analytics for demand forecasting, which Chris Bruman said in Apr 2025 had made forecasts much more accurate and helped Dow optimize inventory; Karen S. Carter said in Jul 2026 that Dow had deployed digital pricing and demand planning capability.

Supply chainIn production
Intelligent pricing and order prediction

Dow uses predictive analytics for intelligent pricing, predicting product delivery issues and missed order opportunities, and its Dow.com B2B platform predicts order delivery timing (Apr 2025).

CommercialIn production
Market Intelligence Hub

Dow's Market Intelligence Hub summarizes news and reports with AI and offers an OpenAI-assisted chatbot over business-curated and external market data, which Dow teams use to shape business strategies and find new opportunities (Mar 2025).

CommercialIn production
Patent review with generative AI

Dow uses generative AI to parse long documents; Chris Bruman said in Apr 2025 that patent reviews for new research fell from months to weeks or even days, and Jim Fitterling said on Jan 29, 2026 that AI in patent research and legal discovery brings big cost savings, without a figure.

R&D and legalIn production
Machine learning in R&D

Dow R&D presented work on polyurethanes predictive intelligence and machine learning models of functional data to a Royal Society of Chemistry webinar in 2025, alongside its Citizen Data Science program, which trains staff outside the central data team to build their own models.

R&DIn production
Application modernization with Kyndryl

Kyndryl is modernizing Dow's infrastructure applications with AI and automation under an expanded engagement announced Nov 10, 2025; a Dow IT director said the collaboration had delivered measurable cost savings, without a figure.

Information TechnologyIn production
Microsoft 365 Copilot for office work

Dow employees use Microsoft 365 Copilot for research and document work, including an innovation specialist who said Copilot shortened hours of data gathering (Nov 2024, Microsoft).

EnterpriseIn production

Adoption

Dow has not itself published how many employees use its AI tools. Microsoft reported that Dow's Microsoft 365 Copilot rollout began with a few hundred employees, reached about 20% of the workforce by Nov 2024 and nearly 60% of the company by Mar 2025,. Chris Bruman said in Apr 2025 that generative AI had brought AI to almost every employee and that Dow had changed its training, security, data privacy and responsible AI policies to match. Since Jan 2026 Dow has described AI as part of Transform to Outperform, but its 2026 earnings calls report progress in roles cut, site playbooks and digital pricing and demand planning rather than AI usage.

By the numbers

Steam cracker furnace operations on C3 AI
50+
C3 AI Reliability, being extended to more furnaces and asset classes (Jun 30, 2025)
Share of Dow on Microsoft 365 Copilot
Nearly 60%
Microsoft blog, Mar 2025; about 20% in Microsoft's Nov 2024 case study; Dow has not published a count
Transform to Outperform target
~$2B
Operating EBITDA uplift by the end of 2027, partly through AI and automation (Oct 2026); no AI share given
Delivered by Sep 2026
~$450M
Transform to Outperform, first 9 months of 2026 (Q3 2026 Form 10-Q); not split by AI
Roles cut under Transform to Outperform
~4,500
Announced Jan 29, 2026; about 70% implemented by Oct 2026
Employees
~31,600
Sep 30, 2026, down from about 34,600 at Dec 31, 2025

Rollout, first pilot to today

  1. Mar 2024

    Dow won a 2024 CIO 100 award for its Integrated Data Hub and paired the hub with a company-wide data and analytics literacy program.

    Dow press release

  2. Nov 2024

    Microsoft reported that Dow's Microsoft 365 Copilot rollout had grown from a few hundred employees to about 20% of the workforce, with 2 supply chain agents under way.

    Microsoft customer story

  3. Mar 2025

    Microsoft reported that Dow had extended Copilot to nearly 60% of the company and had identified more than 100 logistics and operations use cases for agents.

    Microsoft Tech Community

  4. Mar 24, 2025

    Dow won a 2025 CIO 100 award for its Market Intelligence Hub, which includes OpenAI-assisted chat and AI-written summaries.

    Dow press release

  5. Jun 30, 2025

    Univation Technologies, a Dow subsidiary, agreed with C3 AI to license the predictive maintenance software Dow runs on more than 50 steam cracker furnace operations.

    C3 AI and Univation release

  6. Jan 29, 2026

    Dow launched Transform to Outperform, cutting about 4,500 roles and targeting at least $2 billion of Operating EBITDA uplift, in part through AI and automation.

    Form 8-K

  7. Apr 8, 2026

    Dow opened its Delivery Center for AI and analytics at the University of Illinois Research Park.

    University of Illinois Research Park

Economics & controls

Spend and ROI

Technology budget

Dow does not disclose an IT or AI budget. Dow's research and development expense was $752 million in 2025, down from $810 million in 2024 because of cost cuts and lower performance pay (2025 Form 10-K), and Jim Fitterling said on Jan 29, 2026 that some of the money from an earlier $1 billion cost program went into digital capabilities and IT.

What AI itself costs

Dow has not said what it spends on AI. Transform to Outperform carries about $1.1 billion to $1.5 billion of one-time costs, including $600 million to $800 million of severance and $500 million to $700 million of other costs (Jan 2026); Dow had recorded $481 million of severance charges and $194 million of costs to achieve by Sep 30, 2026, and expects about $800 million more cash spending over 2 years (Q3 2026 Form 10-Q). Dow has not said how much of the costs to achieve goes to technology.

Reported ROI

Dow has reported no return it credits to AI alone. Dow delivered about $450 million of Operating EBITDA improvement from Transform to Outperform in the first 9 months of 2026, including about $250 million in the third quarter, expects about $700 million in 2026 and about $2 billion in total by the end of 2027 (the 2026 proxy said by the end of 2028), and says the program works in part through AI and automation. Jim Fitterling said on Jan 29, 2026 that the program is not all AI, and CFO Jeff Tate said in Sep 2026 that the gains come mainly from process optimization, standardization and automation, with AI helping some workflows; Tate's named sources of second-half savings were role cuts, site playbooks, sourcing and contract renegotiation, and logistics optimization. Microsoft reported in Mar 2025 that Dow anticipated millions of dollars of freight savings from its Copilot agents, and Kyndryl quoted a Dow IT director in Nov 2025 on measurable cost savings, but neither gives a realized figure.

Headcount

Dow had about 34,600 employees at Dec 31, 2025 and about 31,600 at Sep 30, 2026. Transform to Outperform cuts about 4,500 Dow roles plus third-party roles, which Dow tied to new ways of working that use automation and AI; about 55% of the cuts were in place by Jul 2026 and about 70% by Oct 2026, with the phase due to be substantially complete by the end of 2026. Dow also cut about 20% of headcount and cost at senior leadership level in the first quarter of 2026 (Apr 2026), and Dow said in Jan 2025 that an earlier $1 billion cost program would cut about 1,500 jobs.

Governance and risk

Board oversight of AI

Dow's Board retains oversight of cybersecurity and artificial intelligence risk, the Audit Committee covers cybersecurity and AI risks within enterprise security, and the Environment, Health, Safety & Technology Committee covers cybersecurity and AI opportunities (2026 proxy). The Audit Committee receives cybersecurity updates at least quarterly and engages with the chief information & digital officer and the chief information security officer, according to the 2025 Form 10-K.

Approved models and data boundary

Chris Bruman said in Apr 2025 that Dow keeps critical data within the company, allows employees to use only pre-approved large language models with data protection, and relies on strategic partners to indemnify it against copyright infringement, and that a Responsible AI framework and principles govern both traditional and generative AI. The Delivery Center counts AI governance among its 4 pillars (Apr 2026).

Disclosed risks

Dow's 2025 Form 10-K says its systems, including third-party systems, may include embedded AI, that attackers increasingly use AI, and that the expanding use of AI in its manufacturing process controls, supply chain optimization and human resources analysis, along with evolving AI regulation, could cause unplanned events that hurt operations or the public.

Recent signals

  1. Oct 22, 2026

    Dow said it had delivered about $450 million of Transform to Outperform benefit in the first 9 months of 2026, about $250 million of it in the third quarter, had implemented about 70% of the 4,500 role cuts, and was using data to tighten its commercial pipeline.

    Q3 2026 earnings call transcript

  2. Sep 2026

    CFO Jeff Tate told a Morgan Stanley conference that Dow's transformation relies mainly on process optimization, standardization and automation, with AI helping some workflows, and that more than 50 process opportunities had been found.

    Constellation Research

  3. Jul 23, 2026

    Karen S. Carter raised Dow's 2026 Transform to Outperform target to about $700 million and cited the deployment of digital pricing and demand planning capability.

    Q2 2026 earnings call transcript

  4. Jul 1, 2026

    Karen S. Carter became CEO of Dow and Jim Fitterling became Executive Chair.

    Form 8-K exhibit

  5. Apr 23, 2026

    Dow said it had begun transformation assessments at about 25% of its large sites and that pilots of digital commercial capabilities had improved the quality and value of new opportunities.

    Q1 2026 earnings call transcript

  6. Apr 8, 2026

    Dow opened a Delivery Center for AI and analytics at the University of Illinois Research Park, inside its Enterprise Data & Analytics organization.

    University of Illinois Research Park

  7. Apr 2026

    Dow won a 2026 CIO 100 award for its Carbon Footprint Ledger, built on Databricks with Boston Consulting Group and integrated into the Integrated Data Hub.

    Chemical Industry Digest

  8. Feb 27, 2026

    Dow's 2026 proxy assigned AI risks to the Audit Committee and AI opportunities to the Environment, Health, Safety & Technology Committee, with the full Board retaining AI oversight.

    2026 proxy statement

  9. Feb 2026

    Dow's 2025 Form 10-K named the expanding use of AI in manufacturing process controls, supply chain optimization and human resources analysis as an operational risk.

    2025 Form 10-K

  10. Jan 29, 2026

    Dow announced Transform to Outperform, cutting about 4,500 roles to add at least $2 billion of Operating EBITDA, in part by using AI and automation; Jim Fitterling said the program is not all AI.

    Form 8-K

  11. Jan 5, 2026

    Dow named Andre Argenton Chief Technology & Sustainability Officer, combining the R&D and sustainability leadership roles.

    2025 Form 10-K

  12. Nov 10, 2025

    Kyndryl expanded its nearly 2-decade collaboration with Dow to modernize Dow's applications with AI and automation.

    Kyndryl

Quick answers

How does Dow deploy AI?
Dow deploys AI through bought tools on its own data: Microsoft 365 Copilot and Copilot Studio agents, C3 AI predictive maintenance on more than 50 steam cracker furnace operations, and an OpenAI-assisted market intelligence hub, all fed by an Integrated Data Hub on Azure and Databricks. Dow tied its Jan 2026 Transform to Outperform plan, about 4,500 role cuts and at least $2 billion of EBITDA uplift, partly to AI and automation, but has not said how much of the roughly $450 million delivered by Sep 2026 came from AI.
What does Dow use AI for?
Dow has described AI use cases in production or pilot from C3 AI predictive maintenance on more than 50 steam cracker furnace operations and Microsoft Copilot agents that audit freight invoices to OpenAI-assisted market intelligence, generative AI for patent review and legal discovery, and predictive models for demand, pricing and delivery. Dow has not published a business return it credits to AI for any of them. At firm level, Dow delivered about $450 million of Operating EBITDA improvement from Transform to Outperform in the first 9 months of 2026 and targets about $2 billion by the end of 2027, but credits that to AI and automation together with process redesign, role cuts and sourcing, with no AI share.
Who leads AI at Dow?
AI at Dow is led by Debra Bauler (Chief Information & Digital Officer); Chris Bruman (Chief Data & Analytics Officer); Mike Weideman (Head of Enterprise Transformation). Ownership of AI on the org chart: Chief Information & Digital Officer Debra Bauler, under CFO Jeff Tate.
What AI platform does Dow use?
Dow buys its AI tools and runs them on its own Integrated Data Hub, built on Microsoft Azure and Databricks, which Chief Data & Analytics Officer Chris Bruman called the biggest change to Dow's AI strategy (Apr 2025).
Which AI models does Dow use?
Dow uses Microsoft 365 Copilot and agents built in Copilot Studio, C3 AI Reliability for equipment, and OpenAI-assisted chat in its Market Intelligence Hub; Bruman said staff may use only pre-approved large language models (Apr 2025).
How many Dow employees use AI?
Microsoft reported in Mar 2025 that Dow had extended Copilot to nearly 60% of the company, up from about 20% in Nov 2024; Dow itself has not published a user count.
What return has Dow reported from AI?
Dow has reported no business return it credits to AI since May 2025. The roughly $450 million of Operating EBITDA improvement Dow delivered from Transform to Outperform in the first 9 months of 2026 (Q3 2026 Form 10-Q) comes from a program the company says works in part through AI and automation, alongside role cuts, site playbooks, sourcing and pricing; Jim Fitterling said on Jan 29, 2026 that it is not all AI, and CFO Jeff Tate said in Sep 2026 that most of it is process optimization and automation. Fitterling's remark that AI brings big cost savings in legal patent and discovery work, and Kyndryl's Nov 2025 statement that their collaboration delivered measurable cost savings, carry no figure. On rollout, Dow has not itself reported how many employees use its AI tools: the nearly 60% Copilot figure comes from a Microsoft blog in Mar 2025, before the window. Dow's AI work is on the recent record, from the C3 AI expansion (Jun 2025) to the Delivery Center for AI and analytics (Apr 2026), so Dow sits at Experimenting rather than Unknown. A return Dow credits to AI in one use case, such as predictive maintenance or freight invoice auditing, or a company-reported AI user count under Chief Information & Digital Officer Debra Bauler, would move Dow to Piloting.

Open questions

  • How much of Transform to Outperform's roughly $450 million delivered by Sep 2026, and of its $2 billion target, Dow credits to AI as opposed to process redesign, role cuts, sourcing and automation.
  • What C3 AI predictive maintenance has returned on Dow's more than 50 steam cracker furnace operations in avoided downtime or maintenance cost; Dow has said only that it anticipates significant benefits.
  • Whether the Copilot freight invoice agents moved past the Mar 2025 proof of concept and what overcharges they recovered against the millions of dollars Dow anticipated.
  • How many Dow employees use Microsoft 365 Copilot or other AI tools today; the only figures, about 20% and nearly 60%, come from Microsoft and predate Nov 2025.
  • Which large language models Dow has pre-approved beyond Copilot and the OpenAI-assisted chat, and where the OpenAI service and C3 AI run.
  • Whether the chief information & digital officer reports to the CFO, as the 2026 proxy implies, or to the chief operating officer, as the 2025 Form 10-K says, after Karen S. Carter moved from COO to CEO.
  • How much Dow spends on IT and AI, and how much of Transform to Outperform's $500 million to $700 million of non-severance one-time costs goes to technology.

Sources

Data confidence

Confirmed
Stated in a primary or reputable public source.
Estimated
Derived or approximated from public reporting.
Unconfirmed
Reported, but not yet corroborated.

Dow profile · Enterprise AI diffusion report by Imagine · facts checked against their sources on Nov 5, 2026