Verified
A.P. Moller - Maersk
Scaling- Shipping and logistics
- 107,638 employees (2025)
- Microsoft Copilot for staff
- ~90 LLM apps on Dify (2025)
- AI risk review per use case
A.P. Moller - Maersk deploys AI through Microsoft Copilot, its preferred generative AI tool for employees, an LLM app platform that vendor Dify says ran about 90 apps with 25,000 active users by late 2025, and operational AI such as Star Connect, which forecasts vessel fuel use. CTIO Navneet Kapoor owns the technology and AI risk, backed by an Enterprise AI Programme, and Chief Data Officer Krishnan Srinivasan presented Maersk's AI North Star to customers in Sep 2026. Maersk reports no financial return from AI.
maersk.comTheir approach to AI
- The platform
- A.P. Moller - Maersk gives employees Microsoft Copilot and, according to vendor Dify, replaced its in-house LLM app platform with Dify Enterprise, which ran about 90 production apps by late 2025.
- The models
- Maersk has not named the models behind its apps; staff use Microsoft Copilot, and Maersk AI engineering job postings ask for Azure OpenAI, Azure AI Foundry and Dify experience.
- The reach
- About 25,000 people actively used Maersk's LLM apps by late 2025, Dify says, citing a Maersk internal assessment; Maersk employed 107,638 people at the end of 2025.
- The catch
- Maersk publishes no AI spending or return figures, its usage numbers come from a vendor, and its CEO named AI in the productivity push behind the Feb 2026 cut of about 1,000 corporate roles, with no figure.
Snapshot
- Employees
- 107,638
- At year-end 2025, in about 130 countries (Annual Report 2025)
- Technology budget
- Not disclosed
- Maersk does not break out technology or AI spending
- Technology workforce
- ~6,000est.
- Estimate: about 2,900 in India said to be almost half of global tech talent (Sep 2025)
- Who owns AI
- Navneet Kapoor, CTIO
- Owner of the technology and AI risk (Annual Report 2025); CDO Krishnan Srinivasan presents the AI North Star
- Platform
- Copilot; Dify Enterprise
- Copilot for staff; LLM apps on Dify, per Dify (late 2025)
- Models
- Not disclosed
- Job postings ask for Azure OpenAI and Azure AI Foundry
- Reach
- ~25,000 active usersest.
- Of Maersk's LLM apps, late 2025, per Dify
- Use cases
- ~90 production appsest.
- Late 2025, per Dify; plus Star Connect and Trade & Tariff Studio
Key findings
Dify says A.P. Moller - Maersk replaced its in-house AI platform with Dify Enterprise after a head-to-head evaluation and had about 90 production LLM apps and 25,000 active users by late 2025, figures Dify attributes to a Maersk internal assessment dated Nov 2025.
Maersk names Microsoft Copilot as the preferred generative AI tool for its employees, used with Maersk credentials, and bars staff from sharing confidential, employee or competitive data and Maersk intellectual property with any AI tool (Maersk Design System, checked Nov 2026).
Star Connect, Maersk's AI-powered fleet energy efficiency platform, runs machine learning at the edge on more than 2.5 billion IoT data points to forecast fuel use, wind resistance and parametric roll risk during voyages (Sep 2025).
Maersk said on Feb 5, 2026 that it would close about 1,000 of roughly 6,000 corporate positions to save $180 million a year, and CEO Vincent Clerc said there would be a very sharp focus on productivity, including AI applications, Bloomberg reported; Maersk did not say how much of the saving AI provides.
Every AI use case at Maersk gets a risk assessment from its application owner, limited-risk systems must show users a pre-approved AI transparency notice, and Maersk's 2025 annual report says it introduced risk-based classification of AI systems that year to align with the EU AI Act.
Where Maersk sits
- Stage
- Scaling
A.P. Moller - Maersk sits at Scaling: Microsoft Copilot is the preferred generative AI tool for Maersk employees, Dify reports about 90 production LLM apps and 25,000 active users at Maersk by late 2025, every AI use case goes through a risk assessment under the Maersk AI Hub, and the 2025 annual report makes CTIO Navneet Kapoor the owner of the technology and AI risk, backed by an Enterprise AI Programme. A.P. Moller - Maersk falls short of Embedded because it reports no financial effect from AI and no weekly usage across its 107,638 staff, and its largest usage figures come from a vendor, not the company.
How the stages are defined - Ownership
- CTIO Navneet Kapoor, owner of the technology and AI risk; CDO Krishnan Srinivasan
- Platform
- Hybrid: Microsoft Copilot bought; LLM apps built on Dify Enterprise
- Models
- Not disclosed; Microsoft Copilot for staff, Azure OpenAI in job postings
- Hosting
- Microsoft Azure for the data platform, Microsoft's cloud for Copilot, edge computers on vessels for Star Connect; Dify hosting not disclosed
- Reach
- About 25,000 active LLM app users (late 2025, per Dify)
How they deploy
Buy the assistant, build the apps
A.P. Moller - Maersk gives employees Microsoft Copilot, a bought product, as its preferred generative AI tool. For business applications, Maersk built an in-house AI platform first and then, according to vendor Dify, replaced it with Dify Enterprise after a controlled head-to-head evaluation; Dify says one app shipped with one person in two weeks against four people in four months on the earlier platform, and that Maersk completed a move to Dify Enterprise v3 in the fourth quarter of 2025.
Educate, Experiment, Execute
Maersk describes its approach to AI in three steps, Educate, Experiment and Execute, set out on the Maersk AI Hub, which also holds the company's AI governance and policy. Maersk's 2025 annual report says the company is committed to driving AI capabilities through an Enterprise AI Programme, and that in 2025 it ran foundational and advanced generative AI courses to build AI fluency across the company.
AI inside the products customers buy
Maersk puts AI into products it sells: Maersk Trade & Tariff Studio for customs classification and tariff planning, expanded in Aug 2026 with a conversational AI interface, Maersk Risk Management for early warnings, and AI investments in Maersk Lead Logistics aimed at visibility, analytics and execution. Chief Data Officer Krishnan Srinivasan presented the Maersk AI North Star, with examples of AI projects available to customers and those coming next, to about 120 customers in Copenhagen in Sep 2026.
India as the engineering base
About 2,900 of Maersk's technology professionals work in Bengaluru, Pune, Mumbai and Chennai, which Resham Sahi, head of the Maersk Technology Centre in Bengaluru, called almost half of the company's global tech talent in Sep 2025. Maersk says the Bengaluru centre builds software, architecture, cybersecurity, data science and AI and holds more than 10 AI patents, and CTIO Navneet Kapoor called India part of the brain and backbone of Maersk's digital transformation.
Productivity under cost pressure
A.P. Moller - Maersk's EBIT fell to $3.5 billion in 2025 from $6.5 billion as freight rates dropped, and on Feb 5, 2026 Maersk announced it would close about 1,000 corporate positions to cut overhead by $180 million a year. CEO Vincent Clerc said there would be a very sharp focus on productivity, including the use of AI applications, Bloomberg reported, without a figure for AI's part.
- Deployment model
- A.P. Moller - Maersk runs technology centrally under CTIO Navneet Kapoor, a member of the Executive Leadership Team. Krishnan Srinivasan, SVP of Technology Services Engineering, whose teams cover compute, integration, network and platform engineering, also holds the Chief Data Officer title and speaks for Maersk's AI plans, and the Maersk Technology Centre in Bengaluru under Resham Sahi builds software, data science and AI. Business units put AI into their products, such as Trade & Tariff Studio in customs consulting, and the Global Service Centres, with more than 14,000 staff at 13 sites in 8 countries, adopt AI use cases in customer service, tenders and disputes that align with the Maersk AI vision. Application owners complete an AI risk assessment for each AI use case.
- Model providers
- A.P. Moller - Maersk has not published which large language models run behind its apps. Microsoft Copilot is the preferred generative AI tool for Maersk employees. Maersk AI engineering job postings ask for experience with Azure OpenAI and Azure AI Foundry, with LLM APIs from OpenAI, Anthropic, Hugging Face, Mistral and Llama, and with orchestration tools such as Dify, which Dify says runs Maersk's production LLM apps. Front-end engineers can add the Maersk Design System's AI agent to Copilot chat in their code editors, and Star Connect runs machine learning models at the edge on vessels.
- Data and infrastructure
- A.P. Moller - Maersk's data teams work on the Azure cloud platform with Databricks and Spark, Dataiku and Data Factory, building data products from a data lake, according to Maersk's careers pages. Maersk Technology Centre hosted its 2020 booking amendment model on Azure as an API, IndiaAI reported in 2021. Star Connect processes more than 2.5 billion IoT data points with machine learning at the edge on board (Sep 2025), and Maersk was completing a new digital connectivity platform on 450 vessels and had fitted next-generation IoT devices to about 30% of its reefers by Jun 2026. Maersk has not said where Dify Enterprise runs or whether any models run in its own data centers.
Platforms and internal tools
Microsoft Copilot
Microsoft Copilot is the preferred natural language generative AI tool for Maersk employees, who sign in with their Maersk credentials and must follow the corporate guidelines of the Maersk AI Hub and the Maersk tone of voice when using it.
- Models
- Microsoft Copilot
LLM app platform on Dify EnterpriseMigration to Dify Enterprise v3 completed Q4 2025
Dify says A.P. Moller - Maersk replaced its in-house AI platform with Dify Enterprise and runs about 90 production apps across customer experience and other areas, led by CX-Assist for retrieval, CX-Email for customer email handling and CX-Tariff Audit for invoice checks. Maersk has not confirmed the figures itself; Dify cites a Maersk internal assessment called Our LLM Journey from Nov 2025.
- Users
- About 25,000 active (late 2025, per Dify)
Star Connect
Star Connect is Maersk's AI-powered fleet energy efficiency platform. Built on edge computing, Star Connect applies machine learning to more than 2.5 billion IoT data points in real time to forecast fuel consumption, wind resistance on cargo and parametric roll risk, so operators can adjust course and speed within minutes.
Maersk Trade & Tariff StudioJun 2025
Maersk Trade & Tariff Studio is a customs and tariff platform with AI-powered tariff engineering across more than 6,000 product codes and 20,000 sub-codes, upstream compliance screening and regulation updates fed by Maersk's 2,700 customs experts. A 2026 version shown at the Maersk and MIT customer forum classifies goods from documents, product links or photos with confidence scores and traceable sources, and the release Maersk announced on Aug 18, 2026 adds a conversational AI interface, AI-assisted pre-entry review and bulk classification of product catalogs.
Port Mirror and Berth Planner (APM Terminals)
APM Terminals, Maersk's port operator, uses AI solutions such as Port Mirror, a digital twin that simulates terminal operations and predicts congestion, and Berth Planner for vessel line-ups and berth optimisation, according to Maersk's 2025 annual report.
Maersk Risk Management
Maersk Risk Management is a digital product that gives customers early warnings of supply chain disruption and recommended actions, which Maersk says is powered by expert insights and AI.
MDS AI Agent and MCP Server
The Maersk Design System's AI agent bundles instructions, skills and component references that engineers install in their code repositories and select in Copilot chat, so generated interface code uses Maersk's own components and design tokens.
Data platform on Azure
Maersk's data engineers ingest data into a data lake and build data products on the Azure cloud platform with Databricks and Spark, Dataiku and Data Factory, and teams choose the technology for their own products.
Vendors and partners
Microsoft
Copilot for employees; Azure cloud for the data platform · Maersk names Microsoft Copilot its preferred generative AI tool for staff, and its data teams build on Azure; Maersk job postings also ask for Azure OpenAI and Azure AI Foundry experience.
DifyBy 2025
LLM application platform, per the vendor · Dify says Maersk chose Dify Enterprise over its in-house platform after a head-to-head evaluation and finished moving to version 3 in the fourth quarter of 2025. Maersk has not announced the relationship itself, but Mark Sear, its Director of AI Solutions Engineering, was quoted on Maersk's use of Dify in Dify's Mar 2026 funding announcement.
Databricks and Dataiku
Data engineering and data science tools · Listed with Spark and Data Factory as the tools Maersk's data teams use (Maersk careers page, checked Nov 2026).
MIT Center for Transportation & Logistics2026
Co-host of Maersk's North American customer forum on AI in supply chains · The forum, Supply Chain of the Future: Methods, Models and What's Next in the Age of AI, is where Maersk showed the 2026 version of Trade & Tariff Studio.
Senseforth.aiJan 2020
Conversational AI start-up in Maersk's OceanPro accelerator · Senseforth.ai joined the third OceanPro cohort to build an AI virtual customer support service for maersk.com over 120 days.
Leadership
AI, data and platform leaders
Executives
AI org map
Central technology and data organization under the CTIO; AI built into business products and adopted in the Global Service Centres
A.P. Moller - Maersk's technology, data and AI work sits under CTIO Navneet Kapoor, a member of the Executive Leadership Team that CEO Vincent Clerc leads and the owner of the technology and AI risk in Maersk's 2025 annual report. Krishnan Srinivasan holds the Chief Data Officer title alongside running Technology Services Engineering, Resham Sahi runs the Maersk Technology Centre in Bengaluru, and Mark Sear is Director of AI Solutions Engineering. Product teams such as Trade & Customs Consulting build AI into what Maersk sells, the Global Service Centres adopt AI in customer service and disputes, and an enterprise Data & AI Ethics Committee governs AI risk.
A.P. Moller - Maersk does not publish an AI or data science headcount. Maersk said in Sep 2025 that about 2,900 technology professionals work in its Indian offices and that this is almost half of its global tech talent, which puts the technology workforce at roughly 6,000 (an estimate). The Technology Services Engineering team under Krishnan Srinivasan has more than 700 members in Bangalore, Pune, Maidenhead, Copenhagen and Mexico and covers database and compute, engineering and operations, enterprise architecture, integration, network and platform engineering, and the Global Service Centres run their own AI and automation team inside an Execution Excellence unit.
Reporting lines
Chief Executive Officer
Technology
Navneet Kapoor, CTIO
Executive Leadership Team member; the line to the CEO is inferred from ELT membership
Technology Services Engineering and Data
Krishnan Srinivasan, SVP and Chief Data Officer
Presented the Maersk AI North Star (Sep 2026); reporting line inferred, not published
Maersk Technology Centre, Bengaluru
Resham Sahi, SVP Technology
About 2,900 technology professionals across India (Sep 2025); reporting line inferred, not published
AI Solutions Engineering
Mark Sear, Director
Named in Dify's Mar 2026 funding announcement; team size and reporting line not published
Global Service Centres
14,000+ colleagues; AI and automation sit in its Execution Excellence unit; reporting line not published
Trade & Customs Consulting
Lars Karlsson
Spoke for Maersk Trade & Tariff Studio at its Jun 2025 launch; reporting line not published
Data & AI Ethics Committee
Enterprise committee with working groups that runs data and AI governance under the executive Risk & Compliance Committee (Annual Report 2025); a Data & AI Ethics team advises on borderline AI risk cases; leader not published
A.P. Moller - Maersk's leadership page, checked Nov 1, 2026, lists Navneet Kapoor as CTIO on the Executive Leadership Team, while the Executive Board now holds only CEO Vincent Clerc and CFO Robert Erni. The same page had not yet shown the new Global Head of Solutions and Chief People Officer that Maersk announced in Aug and Sep 2026, both effective Nov 1, 2026, and Maersk publishes no reporting lines below the Executive Leadership Team.
Where the AI workforce sits
Technology teams in India (Bengaluru, Pune, Mumbai, Chennai)
Estimate of the technology workforce, from Maersk's Sep 2025 statement that about 2,900 Indian technologists are almost half of its global tech talent; Maersk publishes no AI-specific split.
~45-50%Technology teams elsewhere
Estimate: the remainder of the global technology workforce, including Technology Services Engineering teams in other locations.
~50-55%
Named people
Use cases
A.P. Moller - Maersk runs AI in ocean operations, terminals, customer service, customs, software engineering and the products it sells: Star Connect forecasts vessel fuel use, APM Terminals predicts congestion and plans berths with AI, LLM apps answer service agents' questions and handle customer emails and invoice audits, and Trade & Tariff Studio classifies goods for customs. Dify, the vendor behind Maersk's LLM app platform, says about 90 such apps were in production by late 2025. A.P. Moller - Maersk has published no business return for any AI use case and no firm-level AI return; CEO Vincent Clerc named AI as part of the productivity push that came with closing about 1,000 corporate roles in Feb 2026.
| Use case | Area | Status | ROI |
|---|---|---|---|
| Star Connect fleet energy efficiency Star Connect runs machine learning at the edge on vessel IoT data to forecast fuel consumption, wind resistance on cargo and parametric roll risk, so operators can adjust course and speed during voyages. | Ocean | In production | |
| Disruption signals in ocean operations Maersk's ocean operations use AI to pick out the disruption signals that need attention from schedules, terminal productivity and weather data, so teams can prepare recommendations for customers earlier, Maersk's Europe head of ocean operations said in Sep 2026. | Ocean | In production | |
| Congestion forecasting and berth planning at APM Terminals APM Terminals uses Port Mirror, a digital twin that simulates terminal operations and predicts congestion, and Berth Planner for vessel line-ups and berth optimisation; Maersk's 2025 annual report describes APM Terminals Elizabeth, in the Port of New York and New Jersey, using AI platforms this way. | Terminals | In production | |
| CX-Assist for service agents CX-Assist is a single-step retrieval app in Maersk's customer experience organization, built on its Dify platform, that replies in 10 to 15 seconds, according to Dify. | Customer experience | In production | |
| CX-Email customer email handling CX-Email reads and resolves customer emails on Maersk's Dify platform, with a target of resolving 90% of about 20 million emails a year automatically, according to Dify. | Customer experience | In production | |
| CX-Tariff Audit CX-Tariff Audit is an invoice audit app on Maersk's Dify platform that covers 100% of invoices, according to Dify. | Customer experience and finance | In production | |
| Disputes, tenders and sales support The Contract Management & Sales Support team in Maersk's Global Service Centres, about 1,300 people, runs technology and AI-enabled initiatives in long-term agreements, tenders and sales support, including the Athena platform rollout, dispute automation and AI use cases aligned with the Maersk AI vision. | Global Service Centres | In production | |
| Maersk Trade & Tariff Studio Trade & Tariff Studio gives importers AI-powered tariff classification and duty guidance, a conversational interface that takes questions, product photos, documents or links, bulk classification of product catalogs, AI-assisted pre-entry review and continuous scanning of tariff and regulatory changes; Maersk released the expanded version on Aug 18, 2026. | Logistics & Services | In production | |
| Maersk Risk Management Maersk Risk Management gives customers early warnings of disruption and recommended actions, powered by expert insights and AI. | Logistics & Services | In production | |
| AI in Maersk Lead Logistics Maersk says Maersk Lead Logistics has aligned its AI investments with end-to-end visibility, advanced analytics and AI-driven execution for the supply chains it manages for customers. | Logistics & Services | Announced | |
| Design system agent for engineers Maersk front-end engineers can select the Maersk Design System agent in Copilot chat to generate interface code that uses Maersk's own components and design tokens. | Technology | In production | |
| Microsoft Copilot for writing and everyday work Maersk employees use Microsoft Copilot, the company's preferred generative AI tool, for writing and other text work, under the AI Hub's guidelines and a ban on sharing confidential data. | Firmwide | In production |
Adoption
A.P. Moller - Maersk has used machine learning in operations for years: its Bengaluru Technology Centre built an NLP and robotic process automation system for booking amendments in 2020, when Maersk handled about 150,000 booking requests a month, and Star Connect applies machine learning on board its vessels. Generative AI followed on two tracks: Microsoft Copilot as the preferred tool for employees, and an LLM app platform that Dify says carried about 90 production apps and 25,000 active users by late 2025. In 2025 Maersk ran foundational and advanced generative AI courses to build AI fluency across the company, its annual report says. Maersk has published no count of Copilot users, no share of staff using AI weekly and no total of AI use cases in production.
By the numbers
Rollout, first pilot to today
Jan 2020
Maersk's OceanPro start-up accelerator picked Senseforth.ai to build an AI virtual customer support service for maersk.com.
2020
Maersk Technology Centre in Bangalore built a production system that combined NLP models and robotic process automation to process booking amendments, hosted on Azure as an API.
Feb 2020
Navneet Kapoor became Maersk's Chief Technology & Information Officer after three years as Chief Transformation Officer and Head of Global Shared Service Centres.
Apr 2021
Maersk appointed CTIO Navneet Kapoor to its Executive Board, citing the role of technology in the company's transformation.
2025
Maersk ran foundational and advanced generative AI courses to build AI fluency across the company and introduced risk-based classification of AI systems aligned with the EU AI Act, its 2025 annual report says.
Late 2025
Dify says Maersk had about 90 production LLM apps and 25,000 active users on Dify Enterprise and completed a move to version 3 in the fourth quarter of 2025.
Jun 2025
Maersk launched Trade & Tariff Studio, an AI-powered customs and tariff platform, for U.S. imports from Jun 28, 2025 after pilots with large customers.
Jun 2025
Maersk said its Global Service Centres were shaping how the company embeds AI across its operations, in a video with Head of Global Service Centre Caroline Wu.
Sep 2025
Maersk showed Star Connect at Code Cargo 2025 in Bengaluru and said its Indian technology centres hold about 2,900 technologists and more than 10 AI patents.
Aug 2026
Maersk released an expanded Trade & Tariff Studio with a conversational AI interface, AI-assisted pre-entry review and bulk classification of product catalogs.
Sep 2026
Chief Data Officer Krishnan Srinivasan presented the Maersk AI North Star to about 120 customers at the Supply Chain Leaders Forum in Copenhagen.
Economics & controls
Spend and ROI
Technology and AI spending
A.P. Moller - Maersk does not disclose a technology budget or AI spending in its results releases. Maersk reported 2025 revenue of $54.0 billion, EBITDA of $9.5 billion and EBIT of $3.5 billion, down from $6.5 billion in 2024, and guided in Feb 2026 to underlying EBIT of minus $1.5 billion to $1.0 billion for 2026, then raised the range to $2 billion to $4 billion on Jun 29, 2026 and to $4.5 billion to $6.5 billion with its second-quarter results on Aug 13, 2026.
Reported ROI
A.P. Moller - Maersk has not published a business return from AI, for any single use case or for the company. The only figures in circulation come from vendor Dify, citing a Maersk internal assessment from Nov 2025: more than $20 million in estimated annual value, one app shipped 16 times faster on Dify than on the earlier in-house platform, CX-Assist saving more than 90% of the time on single-step lookups, and CX-Email targeting automatic resolution of 90% of about 20 million customer emails a year. None of these is a cost, revenue or headcount figure that Maersk itself has reported.
Headcount and productivity
A.P. Moller - Maersk said on Feb 5, 2026 that it would close about 1,000 of roughly 6,000 corporate positions across headquarters, regions and countries to cut corporate costs by $180 million a year, and CEO Vincent Clerc said there would be a very sharp focus on productivity, including the use of artificial intelligence applications, Bloomberg reported. Maersk did not attribute any part of the $180 million to AI. Maersk's 2025 annual report expects SG&A cost efficiency in Logistics & Services to improve through automation and AI, an assumption in its impairment test rather than a reported result, and puts headcount at 107,638 at the end of 2025.
Governance and risk
AI Hub and policy
A.P. Moller - Maersk keeps its AI governance, compliance rules and corporate AI policy on an internal Maersk AI Hub, and organizes adoption in three steps: Educate, Experiment and Execute. Maersk employees using any AI tool must follow the AI Hub's guidelines, and anyone making images, illustrations, text or video with AI must follow Maersk's AI brand guidelines.
Committees and the EU AI Act
A.P. Moller - Maersk's executive Risk & Compliance Committee oversees its data and AI policies, and an enterprise Data & AI Ethics Committee with supporting working groups runs day-to-day governance, according to the 2025 annual report. In 2025 Maersk introduced risk-based classification of AI systems, documentation practices and oversight routines aligned with the EU AI Act and set up one framework for data and AI risks; the same report rates failure to execute the technology roadmap and adapt to AI advancements as a rising principal risk owned by the CTIO.
Risk assessment for every use case
At Maersk, the application owner completes an AI risk assessment for each AI use case before designers are involved. If the result is limited-risk AI, the application must show users a pre-approved transparency notice, which Maersk links to laws such as the EU AI Act, and teams that want to change the wording or have a borderline case go to the Data & AI Ethics team.
Data boundary for staff
Maersk tells employees never to share confidential information with AI tools, naming business requirements, employee data, competitive strategy, copyrighted material, personal information and Maersk intellectual property. Staff reach Microsoft Copilot with their Maersk credentials.
Human oversight
Maersk's approved notice for internal AI assistants says outputs may be inaccurate, must be reviewed before acting and are not for autonomous decisions, and the notice for customer-facing AI offers a human agent instead. Maersk's conversational AI guidelines require voice systems to disclose AI at the start of a call and again at every handoff.
Cyber security and board oversight
A.P. Moller - Maersk runs cyber security in-house, publishes a cyber security policy and a SOC 3 report on its controls over systems and data. Maersk's board has audit, nomination, remuneration and energy transition committees and no committee dedicated to technology or AI, according to its leadership page checked Nov 1, 2026.
Recent signals
Oct 8, 2026
Maersk reported that Chief Data Officer Krishnan Srinivasan presented the Maersk AI North Star at its AI-themed Supply Chain Leaders Forum in Copenhagen, where 120 customers from more than 70 organizations attended and Vincent Clerc took customer questions.
Sep 29, 2026
Maersk published a Feb 2026 survey of more than 800 European customers finding that 52% use AI in their supply chains and 5% have scaled it, with a video in which Chief Data Officer Krishnan Srinivasan describes the AI projects Maersk is working on.
Aug 18, 2026
Maersk released an expanded Trade & Tariff Studio, its AI-powered customs platform, adding a conversational AI interface, AI-assisted pre-entry review, bulk classification of product catalogs and post-declaration audit support.
Jul 1, 2026
Maersk said Maersk Lead Logistics has aligned its AI investments with visibility, analytics and execution, after a survey in which about 5% of North American customers rated their own logistics AI as advanced.
Jun 26, 2026
Resham Sahi, head of the Maersk Technology Centre in Bengaluru, called generative AI the next frontier for logistics and said it can simplify access to operational intelligence.
May 13, 2026
Maersk published takeaways from its North American customer forum on AI with MIT's Center for Transportation & Logistics, stressing governance, cyber risk and human oversight as conditions for scaling AI.
Mar 9, 2026
Dify's funding announcement named Maersk among the enterprises building on its platform and quoted Mark Sear, Maersk's Director of AI Solutions Engineering, on moving AI workflows from prototype into production.
Feb 5, 2026
A.P. Moller - Maersk's 2025 annual report named failure to execute the technology roadmap and adapt to AI advancements a rising principal risk, owned by the CTIO, and said an Enterprise AI Programme supports Maersk's AI capabilities.
Feb 5, 2026
A.P. Moller - Maersk said it would close about 1,000 of roughly 6,000 corporate positions to save $180 million a year, and CEO Vincent Clerc said there would be a very sharp focus on productivity, including the use of AI applications.
Nov 2025
A Maersk internal assessment, as cited by Dify, counted about 90 production LLM apps and 25,000 active users, and Maersk completed its move to Dify Enterprise v3 in the fourth quarter of 2025.
Quick answers
- How does A.P. Moller - Maersk deploy AI?
- A.P. Moller - Maersk deploys AI through Microsoft Copilot, its preferred generative AI tool for employees, an LLM app platform that vendor Dify says ran about 90 apps with 25,000 active users by late 2025, and operational AI such as Star Connect, which forecasts vessel fuel use. CTIO Navneet Kapoor owns the technology and AI risk, backed by an Enterprise AI Programme, and Chief Data Officer Krishnan Srinivasan presented Maersk's AI North Star to customers in Sep 2026. Maersk reports no financial return from AI.
- What does A.P. Moller - Maersk use AI for?
- A.P. Moller - Maersk runs AI in ocean operations, terminals, customer service, customs, software engineering and the products it sells: Star Connect forecasts vessel fuel use, APM Terminals predicts congestion and plans berths with AI, LLM apps answer service agents' questions and handle customer emails and invoice audits, and Trade & Tariff Studio classifies goods for customs. Dify, the vendor behind Maersk's LLM app platform, says about 90 such apps were in production by late 2025. A.P. Moller - Maersk has published no business return for any AI use case and no firm-level AI return; CEO Vincent Clerc named AI as part of the productivity push that came with closing about 1,000 corporate roles in Feb 2026.
- Who leads AI at A.P. Moller - Maersk?
- AI at A.P. Moller - Maersk is led by Navneet Kapoor (Executive Vice President and Chief Technology & Information Officer); Krishnan Srinivasan (SVP, Technology Services Engineering, and Chief Data Officer); Mark Sear (Director of AI Solutions Engineering (Mar 2026)). Ownership of AI on the org chart: CTIO Navneet Kapoor, owner of the technology and AI risk; CDO Krishnan Srinivasan.
- What AI platform does A.P. Moller - Maersk use?
- A.P. Moller - Maersk gives employees Microsoft Copilot and, according to vendor Dify, replaced its in-house LLM app platform with Dify Enterprise, which ran about 90 production apps by late 2025.
- Which AI models does A.P. Moller - Maersk use?
- Maersk has not named the models behind its apps; staff use Microsoft Copilot, and Maersk AI engineering job postings ask for Azure OpenAI, Azure AI Foundry and Dify experience.
- How many Maersk employees use AI?
- About 25,000 people actively used Maersk's LLM apps by late 2025, Dify says, citing a Maersk internal assessment; Maersk employed 107,638 people at the end of 2025.
- How far has AI spread at A.P. Moller - Maersk?
- A.P. Moller - Maersk sits at Scaling: Microsoft Copilot is the preferred generative AI tool for Maersk employees, Dify reports about 90 production LLM apps and 25,000 active users at Maersk by late 2025, every AI use case goes through a risk assessment under the Maersk AI Hub, and the 2025 annual report makes CTIO Navneet Kapoor the owner of the technology and AI risk, backed by an Enterprise AI Programme. A.P. Moller - Maersk falls short of Embedded because it reports no financial effect from AI and no weekly usage across its 107,638 staff, and its largest usage figures come from a vendor, not the company.
Open questions
- Which large language models power A.P. Moller - Maersk's production apps, and where do they run: Azure OpenAI, other providers, or Maersk's own infrastructure?
- Will Maersk confirm Dify's figures of about 90 production apps, 25,000 active users, about 20 million customer emails a year and more than $20 million in estimated annual value, and what share of those emails is now resolved without a person?
- How many Maersk employees use Microsoft Copilot, and on which license: the free Copilot Chat or paid Microsoft 365 Copilot?
- What business return do Maersk's deployed AI tools deliver, such as fuel saved by Star Connect, cost per customer contact with CX-Email, and revenue from Trade & Tariff Studio?
- How much of the $180 million in corporate cost savings announced in Feb 2026 comes from AI, and which roles did AI replace?
- What do the Maersk AI North Star and the Enterprise AI Programme contain, and what AI results does CTIO Navneet Kapoor, the owner of the technology and AI risk, report to the board?
- How many AI and data science specialists does Maersk employ, and what does it spend on technology and AI each year?
- Who leads Maersk's Data & AI Ethics team, and how many AI use cases have been rated high-risk under the EU AI Act?
Sources
Maersk
- Executive Leadership Team and Board of Directors (checked Nov 1, 2026)
- Navneet Kapoor, CTIO, biography
- Vincent Clerc, CEO, biography
- Maersk appoints CTIO Navneet Kapoor to the Executive Board (Mar 30, 2021)
- Code Cargo 2025: Star Connect and the Bengaluru Technology Centre (Sep 10, 2025)
- Maersk launches Maersk Trade & Tariff Studio (Jun 25, 2025)
- Maersk expands Trade & Tariff Studio to advance customs management (Aug 18, 2026)
- How Maersk Trade & Tariff Studio uses AI in customs management (Aug 18, 2026)
- Supply Chain Leaders Forum 2026: strengthening supply chains through AI (Oct 8, 2026)
- AI's impact on supply chains: company adoption (Sep 29, 2026)
- AI's impact on supply chains: reshaping the global ocean network (Sep 29, 2026)
- AI's impact on supply chains: the evolution of customs compliance (Sep 29, 2026)
- Five takeaways on scaling AI in the supply chain, Maersk and MIT CTL forum (May 13, 2026)
- AI expectations for lead logistics (Jul 1, 2026)
- Maersk deploys next-generation connectivity technology for its reefers (Jul 2, 2026)
- Digital services, including Maersk Risk Management
- Data jobs at Maersk: tools and platforms
- Technology Services Engineering team
- Maersk Global Service Centres
- Cyber security policy and SOC 3 report
- Video: Redefining customer service with AI, Maersk Global Service Centre (Jun 30, 2025)
- Video: Krishnan Srinivasan on AI adoption and Maersk's AI projects (Oct 15, 2026)
Maersk Design System AI guidelines
Results and filings
Partners and vendors
- Dify: logistics and supply chain, Maersk deployment figures (checked Nov 2026)
- Pulse 2.0: Dify raises $30 million, with Mark Sear of Maersk quoted (Mar 9, 2026)
- IT Brief: Dify raises $30 million, Maersk cited as a platform user (Mar 10, 2026)
- Maersk job posting: Senior AI/ML Engineer, Bangalore, via Flexa
- Maersk job posting: Associate AI/ML Engineer, Chengdu, via Flexa
News
- Bloomberg via Transport Topics: Maersk to cut 1,000 jobs, Clerc on productivity and AI (Feb 5, 2026)
- Bloomberg via BGOV: Maersk sees solid demand, capacity overhang risks (Feb 5, 2026)
- India Shipping News: Resham Sahi on technology and AI in logistics (Jun 26, 2026)
- IndiaAI: how Maersk uses AI in on-ground support functions (Sep 13, 2021)
- Business Wire: Senseforth.ai selected for OceanPro, Maersk's start-up accelerator (Jan 28, 2020)
- IoT M2M Council: Maersk picks Senseforth.ai for shipping AI support (Feb 6, 2020)
Data confidence
- Confirmed
- Stated in a primary or reputable public source.
- Estimated
- Derived or approximated from public reporting.
- Unconfirmed
- Reported, but not yet corroborated.
A.P. Moller - Maersk profile · Enterprise AI diffusion report by Imagine · facts checked against their sources on Nov 1, 2026
