Verified
Equinor
Scaling- Energy
- ~24,600 employees
- Omnia data platform on Azure
- Copilot for all staff
- AI saved ~$130M in 2025
- $730M R&D and digital
Equinor deploys AI mainly as traditional machine learning on its own operational data, for predictive maintenance, seismic interpretation and well and field planning. Equinor runs the models from Omnia, its data platform on Microsoft Azure, and its integrated operations centre. Equinor said AI created about $130 million of value and savings for the company and its partners in 2025, and over $330 million since 2020. Hege Skryseth, Executive Vice President for Technology, Digital & Innovation, owns the program, and all employees have Copilot licenses.
equinor.comTheir approach to AI
- The platform
- Equinor builds its AI on Omnia, an enterprise data platform on Microsoft Azure that Equinor has used since 2017, with Cognite Data Fusion as a module. Omnia.Prevent, Equinor's predictive maintenance tool, runs in the integrated operations centre.
- The models
- Equinor says it mainly uses traditional machine learning on its own operational data (Jan 2026) and gives employees copilots, chatbots and agents. Microsoft named Equinor in Oct 2026 among the first users of its Decision-1 model.
- The reach
- Torbjørn Folgerø said in Apr 2026 that Equinor has given Copilot licenses to all employees, around 24,600 people. Equinor has published no count of weekly or active AI users.
- The catch
- Equinor's $130 million AI figure for 2025 covers Equinor and its license partners together, and Equinor has not published its own share or how it measures the value. CFO Torgrim Reitan put the effect at NOK 1 billion or so in Feb 2026 and called it early days.
Snapshot
- Employees
- ~24,600
- End of 2025, in more than 20 countries (2025 annual report)
- Technology budget
- $730M R&D and digital (2025)
- Equinor does not publish a separate IT or AI budget
- AI/ML headcount
- Not disclosedest.
- The Technology, Digital & Innovation business area had 2,129 employees at the end of 2025, across research, IT and digital
- Who owns AI
- Hege Skryseth, EVP Technology, Digital & Innovation
- On the Corporate Executive Committee since Sep 1, 2022
- Platform
- Omnia, on Microsoft Azure
- Enterprise data platform with Cognite Data Fusion as a module; Omnia.Prevent for maintenance
- Models
- Mostly traditional machine learning
- On Equinor's own operational data; copilots, chatbots and agents for staff
- Reach
- Copilot licenses for all employeesest.
- Torbjørn Folgerø to IMD, Apr 2026; no usage count published
- Use cases
- A range in production, 100+ more identified
- Maintenance, seismic, well planning, logistics and trading (Jan and Mar 2026)
Key findings
Equinor said on Jan 7, 2026 that AI contributed about $130 million of value creation and savings for Equinor and its partners in 2025 and over $330 million since 2020. Equinor listed the $130 million as a key figure of its 2025 annual report.
Equinor said in Jan 2026 that predictive maintenance, which monitors more than 700 rotating machines through 24,000 sensors, has created $120 million of value since 2020. Equinor credits AI-driven field planning with a $12 million saving on Johan Sverdrup phase 3.
Equinor's CFO Torgrim Reitan said on the Feb 4, 2026 results call that AI already shows in Equinor's numbers at NOK 1 billion or so. Reitan called these early days and said AI can contribute significantly to further cost improvements.
Equinor says it mainly uses traditional machine learning on its own operational data. Equinor's seismic AI tool raised interpretation capacity 10 times and covered about 2 million square kilometres in 2025.
Equinor named subsurface, operations, supply chain and trading as AI's focus areas at its Jun 16, 2026 Capital Markets Day. Equinor also showed that combining AI agents with physics at Troll could add $400 million of net present value; Equinor has published no result.
Where Equinor sits
- Stage
- Scaling
Equinor reports a firm-level return it credits to AI, which places it at Scaling. Equinor said on Jan 7, 2026 that AI contributed about $130 million of value creation and savings for Equinor and its partners in 2025, and over $330 million since 2020. Equinor repeated the $130 million in its 2025 annual report (Mar 2026), and CEO Anders Opedal said on the Feb 4, 2026 results call that AI saved $130 million in 2025 and that the effect is accelerating. Equinor also credits AI with 2 use-case returns in different areas: $120 million of value from predictive maintenance since 2020, and a $12 million saving for the Johan Sverdrup phase 3 partnership from AI-driven field planning (Jan 2026). On rollout, Equinor runs AI on its Omnia data platform on Microsoft Azure, and Executive Vice President Hege Skryseth owns the program and reports its results. Equinor falls short of Embedded because it has published no count of employees who use AI each week, only that every employee has a Copilot license (Apr 2026). Equinor also falls short because the $130 million covers its license partners and is small against Equinor's cost base.
How the stages are defined - Ownership
- Hege Skryseth, EVP Technology, Digital & Innovation, on the Corporate Executive Committee
- Platform
- Hybrid: Omnia data platform built on Azure; Copilot and Cognite bought
- Models
- Mostly in-house machine learning; Copilot and Microsoft models for staff
- Hosting
- Microsoft Azure (Omnia, since 2017) and Databricks for a seismic pipeline; large language model hosting not disclosed
- Reach
- Copilot licenses for all ~24,600 employees (Apr 2026)
How they deploy
Industrial machine learning first
Equinor says it primarily uses traditional machine learning on its own operational data, and that AI now solves industrial tasks on its offshore platforms and onshore plants (Jan 2026). Equinor's 2025 annual report says the company is embedding AI across the value chain, from exploration and subsurface to operations, trading and administration. The annual report says 2025 brought accelerated adoption of industrial AI with tangible value.
AI tied to the Norwegian shelf
Equinor calls AI crucial to its ambition of holding production on the Norwegian continental shelf at 2020 levels, around 1.2 million barrels of oil equivalent a day, through 2035. Hege Skryseth said Equinor uses AI to interpret more seismic data, plan and drill more wells and run its facilities, and to cut energy use and CO2 emissions (Jan 2026).
Data platform as the base
Equinor built Omnia, its enterprise data platform, on Microsoft Azure, which it began adopting in 2017. Equinor added Cognite Data Fusion as a module in Omnia under a long-term frame agreement in Nov 2022. Equinor also uses Microsoft Azure Data Manager for Energy, an OSDU platform, as another part of Omnia for subsurface data.
Generative AI for staff
Equinor gives employees copilots, chatbots and agents to solve tasks and work in new ways, Hege Skryseth said in Jan 2026. Torbjørn Folgerø told IMD in Apr 2026 that Equinor has given Copilot licenses to all employees. Folgerø said tools alone are not enough and that processes such as drilling, maintenance and finance need redesign with attention from the top. Equinor's 2025 annual report says it aims to use generative AI on industrial use cases in operations and maintenance, supply chain and trading.
Focused technology spending
Equinor told investors at its Jun 16, 2026 Capital Markets Day that technology development targets the highest-value opportunities and that AI is an integrated part of everything it does. Equinor named AI's focus areas as subsurface, operations, supply chain and trading, across the international oil and gas and power businesses.
- Deployment model
- Equinor runs AI from its Technology, Digital & Innovation (TDI) business area, led by Executive Vice President Hege Skryseth. TDI brings together research, technology development, digital, IT, innovation and ventures. TDI's Technology & Improvements area has clusters for Enterprise Digital, Oil and Gas, Renewables and Low Carbon, and Technology Strategy and Portfolio and Partnerships (2025 annual report). Equinor's business areas run the tools in daily work. The integrated operations centre uses Omnia.Prevent and automatic process control across assets, and Marketing, Midstream & Processing uses AI to optimise cargo, bunkering and routing in trading (Capital Markets Day, Jun 2026).
- Model providers
- Equinor says most of its AI value comes from traditional machine learning on its own operational data. Omnia.Prevent alone ran more than 520 machine learning models in production, according to an Equinor page on the tool. For office work, Equinor gives employees chatbots and agents (Jan 2026) and Copilot, which Torbjørn Folgerø said in Apr 2026 every employee has. Microsoft named Equinor on Oct 28, 2026 among more than 4,000 customers that used its Decision-1 model, built for structured decision-making, in the model's first month. Equinor has not said which large language models run behind its chatbots and agents.
- Data and infrastructure
- Equinor runs its enterprise data platform, Omnia, on Microsoft Azure, which it began adopting in 2017. Microsoft Azure Data Manager for Energy, an OSDU data platform, is also part of Omnia (Microsoft customer story). Cognite Data Fusion is a module in Omnia under a Nov 2022 frame agreement. Equinor rebuilt a 4D seismic interpretation pipeline on the Databricks Data + AI Platform with Ray and Apache Spark under Unity Catalog (Databricks, May 2025). Databricks did not say which cloud the pipeline runs on. Equinor streams data to its integrated operations centre, whose database held 85 billion data points (Capital Markets Day, Jun 2026). Equinor has not said where its generative AI models run or whether it runs models in its own data centers.
Platforms and internal tools
OmniaAzure adoption from 2017
Omnia is Equinor's enterprise data platform and data architecture on Microsoft Azure, which brings Equinor's operational and subsurface data together for analytics and AI. Cognite Data Fusion and Microsoft Azure Data Manager for Energy run as modules of Omnia.
- Models
- Machine learning models built by Equinor teams
Omnia.Prevent
Omnia.Prevent is Equinor's machine learning maintenance tool for its integrated operations centre. The tool's models watch sensor data from turbines, compressors and generators and flag anomalies, and an engineer reviews each alarm before the centre is notified. Equinor said in Jan 2026 that predictive maintenance covers more than 700 rotating machines and 24,000 sensors across all facilities.
- Users
- Equinor's integrated operations centre and maintenance engineers
- Models
- More than 520 machine learning models in production (Equinor page, undated)
Integrated operations centre
Equinor's integrated operations centre monitors data across assets, using automatic process control to optimise production and condition monitoring to reduce downtime and unplanned shutdowns. Equinor said 85 billion data points stream to the centre's database (Capital Markets Day, Jun 2026).
AI seismic interpretation tool
Equinor's AI tool for seismic interpretation increased interpretation capacity 10 times and was used on about 2 million square kilometres of seismic data in 2025. The work is part of Equinor's rescanning of the Norwegian continental shelf with new and old seismic.
GATE gas optimisation engine
GATE is the gas optimisation engine of Equinor's trading business and applies advanced analytics to data and market insights. Equinor presented GATE in Jun 2026 alongside AI-driven optimisation of cargo, bunkering and routing, more algorithmic trading and back-office automation.
Vendors and partners
Microsoft2017
Azure cloud for the Omnia data platform, Azure Data Manager for Energy, Copilot and the Decision-1 model · Microsoft named Equinor among the first-month users of its Decision-1 model on Oct 28, 2026.
CogniteDec 2021, extended company-wide Nov 2022
Cognite Data Fusion as a module in Omnia, for industrial data extraction and contextualization
DatabricksMay 2025
Data + AI Platform for a 4D seismic interpretation pipeline
DNVApr 2026
Convenor of a Norwegian regulatory sandbox for high-risk industrial AI · Equinor signed the Jan 2026 NHO technology declaration from which the DNV consortium grew.
Leadership
AI, data and platform leaders
Executives
AI org map
Central technology and digital business area on the executive committee, AI used in the operating business areas
Equinor puts AI inside Technology, Digital & Innovation (TDI), one of its business areas, led by Executive Vice President Hege Skryseth on the Corporate Executive Committee. TDI's Technology & Improvements area holds the Enterprise Digital, Oil and Gas, Renewables and Low Carbon, and Technology Strategy and Portfolio and Partnerships clusters, and New Business & Investments runs ventures. Equinor's operating business areas use the tools: Exploration & Production Norway in operations, Projects, Drilling & Procurement in field development, and Marketing, Midstream & Processing in trading.
Equinor has not published the size of its AI and data science workforce. Equinor's Technology, Digital & Innovation business area, which houses research, digital, IT and innovation, had 2,129 employees at the end of 2025 (2025 annual report). Equinor's AI teams date to 2015, when Torbjørn Folgerø set up the company's first data science team. Omnia.Prevent is run by data scientists, developers and engineers who review each model alarm before the integrated operations centre acts on it.
Reporting lines
President and CEO
Anders Opedal
Technology, Digital & Innovation
Hege Skryseth, Executive Vice President
Research, technology, digital, IT, innovation and ventures; 2,129 employees at the end of 2025
Technology & Improvements
Clusters for Enterprise Digital, Oil and Gas, Renewables and Low Carbon, and Technology Strategy and Portfolio and Partnerships
Digital and AI transformation
Torbjørn Folgerø
Placement and title not confirmed; IMD describes Folgerø as leading Equinor's digital and AI transformation (Apr 2026)
New Business & Investments
New businesses outside the core and venture investments
Exploration & Production Norway
Kjetil Hove, Executive Vice President
About two-thirds of equity production; seismic AI and predictive maintenance in use
Projects, Drilling & Procurement
Geir Tungesvik, Executive Vice President
Field development and wells; AI-driven well and field planning
Marketing, Midstream & Processing
Irene Rummelhoff, Executive Vice President
Trading; GATE gas optimisation engine and AI-driven cargo routing
Integrated operations centre
Runs Omnia.Prevent, automatic process control and condition monitoring across assets; reporting line not published
Equinor's executive committee page and 2025 annual report confirm the business areas and their leaders. Equinor has not published where the integrated operations centre reports, nor Torbjørn Folgerø's current title and reporting line.
Where the AI workforce sits
Technology, Digital & Innovation (digital, data science, IT)
Estimate: Equinor does not say how many of TDI's 2,129 employees work on AI.
~50-60%Operations and subsurface teams in the business areas
Estimate: engineers and data scientists in the integrated operations centre, maintenance and exploration who run the models.
~25-35%Trading and supply chain analytics
Estimate: Marketing, Midstream & Processing teams behind GATE and cargo optimisation.
~10-15%
Named people
Use cases
Equinor says it has a range of AI solutions in use and has identified more than 100 new use cases (Jan 2026). Equinor's AI uses span predictive maintenance, seismic interpretation, well and field planning, logistics, trading, wind farm maintenance, environmental monitoring and office work. Equinor credits AI with returns in 2 of them: $120 million from predictive maintenance since 2020 and $12 million saved on Johan Sverdrup phase 3. Equinor also reports a firm-level figure: AI contributed about $130 million of value and savings for Equinor and its partners in 2025, and over $330 million since 2020.
| Use case | Area | Status | ROI |
|---|---|---|---|
| Predictive maintenance of rotating equipment (Omnia.Prevent) Machine learning models monitor more than 700 rotating machines through 24,000 sensors across all Equinor facilities and predict failures and maintenance needs. The models cut the risk of sudden shutdowns that can lead to flaring, and an engineer reviews each alarm before the integrated operations centre acts. | Operations and maintenance | In production | $120M of value since 2020 (Jan 2026) |
| AI-driven well and field development planning AI generates thousands of alternatives for wells and field developments so experts can focus on the best proposals; at Johan Sverdrup phase 3 the tool found an option no one had considered. | Field development | In production | $12M saved for the Johan Sverdrup phase 3 partnership (Jan 2026) |
| AI seismic interpretation Equinor's AI tool raises seismic interpretation capacity 10 times and covered about 2 million square kilometres in 2025. Anders Opedal said in Feb 2026 that new seismic and AI helped find discoveries such as Lofn and Langemann in the Sleipner area. In Jun 2026 Equinor counted 27 discoveries since 2024, without saying how many came from AI. | Exploration | In production | |
| 4D seismic pipeline on Databricks Equinor rebuilt one of its 4D seismic interpretation workflows as a distributed pipeline on Databricks. The workflow detects and classifies reservoir changes over time to pick drilling targets, and Databricks reported cost cuts of up to 96% on specific calculation steps (May 2025). | Exploration | In production | |
| Integrated operations centre process control and condition monitoring Equinor's integrated operations centre uses automatic process control and condition monitoring across assets. Equinor put the centre's realised cash flow impact in 2025 at $3 billion (100%, pre-tax) at its Jun 2026 Capital Markets Day. The figure covers the whole centre, and Equinor does not credit it to AI alone. | Operations and maintenance | In production | |
| AI agents with physics models at Troll Equinor showed at its Jun 2026 Capital Markets Day that combining AI agents and physics at the Troll field could add $400 million of net present value (100%, pre-tax). Equinor has published no realised result. | Operations and maintenance | Announced | |
| Shipping capacity and route optimisation Equinor uses analytics and AI in logistics and operational planning to optimise shipping capacity use and routes on the Norwegian continental shelf (2025 annual report). | Logistics | In production | |
| Cargo, bunkering and routing optimisation Equinor's trading business uses AI to optimise cargo, bunkering and routing beside GATE, its gas optimisation engine. Equinor said GATE earned $60 million of adjusted operating income in 2025 from advanced analytics, a figure it does not credit to AI (Jun 2026). | Trading | In production | |
| Condition-based maintenance for wind farms Equinor's 2025 annual report says AI for condition-based maintenance of wind farms has reduced unplanned downtime through better maintenance plans and insight into component condition. Equinor gives no figure. | Power | In production | |
| Seabed biology and bird detection Equinor uses AI and machine learning to detect seabed biology in video from its underwater drones and to detect birds near wind turbines and oil and gas installations to monitor collision risk (2025 annual report). | Environment | In production | |
| Copilot, chatbots and agents for employees Equinor employees can use copilots, chatbots and AI agents to solve tasks (Jan 2026). Torbjørn Folgerø said in Apr 2026 that every employee has a Copilot license. Equinor has not published usage or which tasks the agents handle. | Firmwide | In production |
Adoption
Equinor's AI spread in 2 layers. Machine learning on operational data came first: Torbjørn Folgerø set up Equinor's first data science team in 2015, and Equinor began adopting Azure in 2017. The value Equinor counts from AI starts in 2020, and Omnia.Prevent detected more than 200 failures on heavy rotating equipment in 2022 alone. Generative AI for office work followed: Equinor says employees can use copilots, chatbots and agents (Jan 2026), and Folgerø said in Apr 2026 that all employees have Copilot licenses. Equinor has published no count of weekly or active AI users among its around 24,600 employees.
By the numbers
Rollout, first pilot to today
2015
Torbjørn Folgerø set up Equinor's first data science team.
2017
Equinor began adopting Microsoft Azure, the cloud under its Omnia data platform.
2017
Equinor created a Chief Digital Officer role and a digital innovation centre, and Torbjørn Folgerø took the role.
2022
Omnia.Prevent, Equinor's machine learning maintenance tool, detected more than 200 failures on heavy rotating equipment during the year.
Sep 2022
Hege Skryseth joined Equinor as Executive Vice President for Technology, Digital & Innovation from Kongsberg Digital.
Nov 2022
Equinor signed a long-term frame agreement with Cognite to use Cognite Data Fusion as a module in its Omnia data architecture on Azure.
May 2025
Databricks described Equinor's rebuilt 4D seismic interpretation pipeline on the Databricks Data + AI Platform.
Jan 2026
Equinor reported that AI created about $130 million of value and savings in 2025 and over $330 million since 2020, and that employees can use copilots, chatbots and agents.
Apr 2026
Torbjørn Folgerø said Equinor had given Copilot licenses to all employees.
Jun 2026
Equinor named subsurface, operations, supply chain and trading as AI's focus areas at its Capital Markets Day.
Economics & controls
Spend and ROI
Technology budget
Equinor invested $730 million in research and development and digital in 2025, a level it said it had stabilised (2025 annual report). Equinor does not publish a separate IT or AI budget.
What AI itself costs
Equinor has not disclosed what it spends on AI, its Microsoft and Copilot licenses, or the Cognite and Databricks platforms. Equinor said at its Jun 2026 Capital Markets Day that investments in new technologies return 3 times their cost in cash flow improvements, a figure for technology as a whole, not AI.
Reported ROI
Equinor said on Jan 7, 2026 that AI contributed about $130 million of value creation and savings for Equinor and its partners in 2025. Hege Skryseth said Equinor had realised over $330 million with AI in industrial processes since 2020, and the 2025 annual report lists the $130 million as approximate value created from AI. On the Feb 4, 2026 results call, Anders Opedal said AI saved $130 million in 2025 and that the effect is accelerating. Torgrim Reitan put the effect in Equinor's numbers at NOK 1 billion or so. Equinor has not said what share of the $130 million falls to Equinor rather than its license partners, or how much is cost saved against value created.
Figures not credited to AI
Equinor's Jun 2026 Capital Markets Day put the integrated operations centre's realised cash flow impact in 2025 at $3 billion (100%, pre-tax), from process control and condition monitoring together. The same presentation put GATE, the trading business's gas optimisation engine, at $60 million of adjusted operating income in 2025. Equinor also targets $100 million of additional adjusted operating income from digital trading solutions by 2030 against 2025, and showed that AI agents and physics at Troll could add $400 million of net present value. Equinor credits none of these to AI alone.
Headcount and costs
Equinor has made no statement linking headcount to AI. Torgrim Reitan said on Feb 4, 2026 that Equinor expects reported operating and administrative costs to fall 10% in 2026, mostly from divestments and structural changes. Reitan named AI and the NCS 2035 work on the Norwegian continental shelf as future sources of cost improvement.
Governance and risk
Risk-based approach
Equinor's 2025 annual report says the company takes a risk-based strategy toward AI, stressing safe and responsible use, and keeps employees involved through targeted upskilling programmes. Equinor applies the EU AI Act's definition of an AI system across its operations, while abiding by local definitions where regulations differ.
Engineers in the loop
Omnia.Prevent's model alarms go first to an engineer on the Prevent team, who decides whether the behaviour is normal or whether to notify the integrated operations centre. The centre then decides whether to investigate (Equinor). Torbjørn Folgerø told IMD in Apr 2026 that Equinor is both optimistic and cautious about generative and agentic AI and aims to adopt it safely and cost-effectively.
Disclosed risks
Equinor's 2025 annual report says AI is complex and uncertain and brings risks in cybersecurity, data privacy, inaccuracies, bias or discrimination and intellectual property infringement. The report says AI may compound cyber risks, and that failure to deploy new technologies well could bring legal action and reputational harm.
Regulator scrutiny of digital drilling
Norway's offshore safety regulator Havtil issued an order to Equinor on Dec 19, 2025 after auditing digital well planning, automated drilling and digital twins on the Askepott rig. Havtil found the Novos automated drilling technology installed or planned on several mobile rigs while Equinor kept its qualification at readiness level 4, called that use without qualification a serious breach, and ordered Equinor to show Novos meets the regulations before adopting it on further offshore facilities (Offshore, Jan 8, 2026).
Industrial AI sandbox
Equinor signed a Jan 2026 technology declaration by NHO and Norwegian industrial companies. DNV announced on Apr 27, 2026 a national consortium, grown from that declaration, to build a regulatory sandbox for high-risk industrial AI, such as systems that influence or control industrial assets.
Recent signals
Oct 28, 2026
Microsoft named Equinor on its fiscal first-quarter call among more than 4,000 customers that used its new Decision-1 model for structured decision-making in its first month.
Jun 16, 2026
At its Capital Markets Day, Equinor called AI an integrated part of everything it does and named subsurface, operations, supply chain and trading as AI's focus areas. Equinor also showed that AI agents and physics at Troll could add $400 million of net present value.
Jun 16, 2026
Equinor said it aims to capture more trading value from its flexible portfolio and from advancing digital tools and AI. Equinor expects adjusted operating income from trading and market optimisation to rise 25% to around $500 million a quarter by 2030.
Apr 30, 2026
Torbjørn Folgerø told IMD that Equinor has given Copilot licenses to all employees and that accelerating responsible AI adoption is a key focus.
Apr 27, 2026
DNV announced a national consortium to build a regulatory sandbox for high-risk industrial AI in Norway, growing from a Jan 2026 NHO declaration that Equinor signed.
Mar 19, 2026
Equinor's 2025 annual report listed about $130 million of value created from AI and $730 million invested in R&D and digital in 2025.
Feb 4, 2026
On the 2025 results call Anders Opedal said AI saved $130 million in 2025 and is accelerating, and Torgrim Reitan said AI shows in Equinor's numbers at NOK 1 billion or so.
Jan 7, 2026
Equinor said AI contributed about $130 million of value and savings for Equinor and its partners in 2025. Equinor credited predictive maintenance with $120 million of value since 2020 and AI-driven field planning with $12 million saved at Johan Sverdrup phase 3.
Dec 19, 2025
Havtil, which had found the Novos automated drilling technology installed or planned on several mobile rigs without completed qualification and called that a serious breach, ordered Equinor to show Novos meets Norwegian requirements before adopting it on further offshore facilities.
Quick answers
- How does Equinor deploy AI?
- Equinor deploys AI mainly as traditional machine learning on its own operational data, for predictive maintenance, seismic interpretation and well and field planning. Equinor runs the models from Omnia, its data platform on Microsoft Azure, and its integrated operations centre. Equinor said AI created about $130 million of value and savings for the company and its partners in 2025, and over $330 million since 2020. Hege Skryseth, Executive Vice President for Technology, Digital & Innovation, owns the program, and all employees have Copilot licenses.
- What does Equinor use AI for?
- Equinor says it has a range of AI solutions in use and has identified more than 100 new use cases (Jan 2026). Equinor's AI uses span predictive maintenance, seismic interpretation, well and field planning, logistics, trading, wind farm maintenance, environmental monitoring and office work. Equinor credits AI with returns in 2 of them: $120 million from predictive maintenance since 2020 and $12 million saved on Johan Sverdrup phase 3. Equinor also reports a firm-level figure: AI contributed about $130 million of value and savings for Equinor and its partners in 2025, and over $330 million since 2020.
- Who leads AI at Equinor?
- AI at Equinor is led by Hege Skryseth (Executive Vice President, Technology, Digital & Innovation); Torbjørn Folgerø (Leads Equinor's digital and AI transformation). Ownership of AI on the org chart: Hege Skryseth, EVP Technology, Digital & Innovation, on the Corporate Executive Committee.
- What AI platform does Equinor use?
- Equinor builds its AI on Omnia, an enterprise data platform on Microsoft Azure that Equinor has used since 2017, with Cognite Data Fusion as a module. Omnia.Prevent, Equinor's predictive maintenance tool, runs in the integrated operations centre.
- Which AI models does Equinor use?
- Equinor says it mainly uses traditional machine learning on its own operational data (Jan 2026) and gives employees copilots, chatbots and agents. Microsoft named Equinor in Oct 2026 among the first users of its Decision-1 model.
- How many Equinor employees use AI?
- Torbjørn Folgerø said in Apr 2026 that Equinor has given Copilot licenses to all employees, around 24,600 people. Equinor has published no count of weekly or active AI users.
- What return has Equinor reported from AI?
- Equinor reports a firm-level return it credits to AI, which places it at Scaling. Equinor said on Jan 7, 2026 that AI contributed about $130 million of value creation and savings for Equinor and its partners in 2025, and over $330 million since 2020. Equinor repeated the $130 million in its 2025 annual report (Mar 2026), and CEO Anders Opedal said on the Feb 4, 2026 results call that AI saved $130 million in 2025 and that the effect is accelerating. Equinor also credits AI with 2 use-case returns in different areas: $120 million of value from predictive maintenance since 2020, and a $12 million saving for the Johan Sverdrup phase 3 partnership from AI-driven field planning (Jan 2026). On rollout, Equinor runs AI on its Omnia data platform on Microsoft Azure, and Executive Vice President Hege Skryseth owns the program and reports its results. Equinor falls short of Embedded because it has published no count of employees who use AI each week, only that every employee has a Copilot license (Apr 2026). Equinor also falls short because the $130 million covers its license partners and is small against Equinor's cost base.
Open questions
- What share of the $130 million of AI value for 2025 accrues to Equinor rather than its license partners, and how much is cash cost saved versus value created?
- How does Equinor measure AI value, and why did CFO Torgrim Reitan cite NOK 1 billion or so in Feb 2026 when the Jan 2026 release gave NOK 1.3 billion?
- How many Equinor employees use Copilot, chatbots or agents each week, and which agents are in production?
- Which large language models run behind Equinor's chatbots and agents, and where are they hosted?
- What return, if any, has the seismic interpretation tool produced in discoveries, drilling or cost, beyond its 10 times capacity gain?
- What do the logistics, trading, wind farm maintenance and environmental monitoring uses of AI return?
- Has the AI agents and physics work at Troll moved from a $400 million NPV potential to realised production?
- What is Torbjørn Folgerø's current title, and where does the integrated operations centre sit in Equinor's organization?
Sources
Equinor
- Use of artificial intelligence saved Equinor USD 130 million in 2025 (Jan 7, 2026)
- Kunstig intelligens ga 1,3 milliarder i besparelser, Norwegian release (Jan 7, 2026)
- Equinor's Capital Markets Day 2026, release (Jun 16, 2026)
- Corporate executive committee (checked Nov 2026)
- Omnia.Prevent: machine learning for maintenance
Filings, results and investor materials
Partners
- Microsoft FY27 first quarter earnings call (Oct 28, 2026)
- Microsoft customer story: Equinor and Azure Data Manager for Energy
- Cognite: Equinor and Cognite enter long-term collaboration (Nov 22, 2022)
- Databricks: How Equinor optimized its seismic data pipeline (May 12, 2025)
- DNV: National consortium to accelerate safe industrial AI (Apr 27, 2026)
News
- JPT: Equinor says AI saved it $130M in 2025 (Jan 2026)
- Offshore: AI helping Equinor speed up seismic interpretation, well planning (Jan 2026)
- OE Digital: Equinor's AI drive yields $130M in savings (Jan 2026)
- Riviera: AI has saved Equinor US$330M since 2020 (Jan 2026)
- IMD: The engineer who is helping power Europe's energy future (Apr 30, 2026)
- Energy Council: Torbjørn Folgerø, SVP and Chief Digital Officer (2018)
Data confidence
- Confirmed
- Stated in a primary or reputable public source.
- Estimated
- Derived or approximated from public reporting.
- Unconfirmed
- Reported, but not yet corroborated.
Equinor profile · Enterprise AI diffusion report by Imagine · facts checked against their sources on Nov 9, 2026
