Marathon Petroleum AI Adoption Tracker
Last updated: September 18, 2026
Overview
Marathon Petroleum Corporation has positioned itself as a leader in digital refinery transformation, leveraging AI technologies for operational excellence and future-proofing its business model[1][2]. The company has deployed AI across multiple operational functions, including predictive maintenance, refinery optimization, and coker cycle management, with notable successes such as a 25% reduction in suboptimal drum cycles in their coke drum operations[3]. Marathon has also embraced AI-driven partnerships and strategic initiatives, including a collaboration with MPLX and MARA Holdings to power AI data centers with natural gas, representing a strategic pivot from internal AI optimization to becoming an external enabler of the AI ecosystem[4][5].
- [1] Marathon's 2025 AI Strategy: Powering Data Centers
- [2] Marathon's 2025 AI Power Play: Fueling Data Centers
- [3] Marathon, CPChem, Big West leverage AI to boost speed and reliability
- [4] MPLX and MARA Announce Collaboration on Integrated Power Generation and Data Center Campuses in West Texas
- [5] Opportunities for Positive ROI from the Application of AI to Refinery Processes
AI Maturity Index
Radar Comparison
Peer Comparison: Marathon Petroleum vs energy
Based on 22 companies in sector
| Dimension | Marathon Petroleum | Sector Avg | Diff |
|---|---|---|---|
| Adoption | 3.0 | 3.3 | -0.3 |
| Proficiency | 2.0 | 2.9 | -0.9 |
| Impact | 4.0 | 3.4 | +0.6 |
| Overall | 3.1 | 3.2 | -0.1 |
AI Hiring Signals
Marathon Petroleum Job Postings Analysis
Tech vs Non-Tech AI Requirements
Non-tech AI adoption exceeds tech roles - strong org-wide AI fluency signal
Top Departments by AI Mention Rate
Analysis
AI hiring signals are limited at Marathon Petroleum, with only 3.8% of job postings mentioning AI. Interestingly, non-tech roles show higher AI requirements (5.0%) than tech roles (0.0%), suggesting AI adoption may be more focused on business applications than technical development. Customer service leads with 25% AI mention rate, indicating operational AI deployment priorities.
View Sample Job Postings (6 sources)
Key Metrics
AI Initiatives
Square Robot Submersible Tank Inspection
December 2025
Partnership for next-generation robotic tank inspection platform
AI-powered submersible robots for in-service tank inspections to enhance safety and reduce downtime
Proprietary Large Language Model
November 2025
Company-specific version of ChatGPT for technical standards
Makes it easier to find and verify technical standards, with direct references to corporate documents
MPLX-MARA AI Data Center Collaboration
November 2025
Strategic partnership to provide natural gas for AI data center power generation
1.5-gigawatt power deal in West Texas to support AI workloads and high-performance computing
Imubit AI Process Optimization
March 2025
AI-powered closed-loop process optimization for refinery operations
Achieved 25% reduction in suboptimal drum cycles and significant improvement in coker giveaway
AI-Ready Operations Maintenance Transformation
2025
Standardization of maintenance operations across midstream business to enable AI
Building digital and data infrastructure for predictive maintenance, intelligent scheduling, and frontline decision-making
Dynamic Risk Analyzer (DRA) for Anomaly Detection
2025
Hands-off AI solution for predictive maintenance in refinery operations
Automates data cleansing, model development, and real-time monitoring of thousands of process signals
C3 AI Reliability Platform
2021-2024
Deployed across 16 refineries for predictive maintenance
Expected to generate $50 million in annual economic benefits
Frequently Asked Questions
Marathon uses C3 AI Reliability for predictive maintenance across 16 refineries, Imubit for closed-loop process optimization, Dynamic Risk Analyzer for anomaly detection, and a proprietary large language model for technical standards verification.
Through its subsidiary MPLX, Marathon has partnered with MARA Holdings to provide natural gas for AI data center power generation in West Texas, with a 1.5-gigawatt power deal to support high-performance computing and AI workloads.
Marathon has achieved a 25% reduction in suboptimal drum cycles in coker operations, expects $50 million in annual economic benefits from predictive maintenance, and has improved operational efficiency through AI-driven process optimization.
Marathon uses a hybrid approach, partnering with AI vendors like C3 AI and Imubit while also developing internal capabilities such as a proprietary large language model for technical standards.
AI enables predictive maintenance to prevent equipment failures before they occur, automates data analysis across thousands of process signals, and helps transition from reactive to predictive maintenance strategies, reducing downtime and costs.
In Application
| Application | Vendor | Use Case |
|---|---|---|
| C3 AI Reliability | C3 AI | Predictive maintenance across 16 refineries on Microsoft Azure platform |
| Imubit Process Optimization | Imubit | Closed-loop refinery process optimization and margin enhancement |
| Dynamic Risk Analyzer (DRA) | Near-Miss Management | Autonomous anomaly detection and predictive maintenance for refinery operations |
| Proprietary LLM | Internal development | Technical standards search and verification with document referencing |
Sources
Marathon's 2025 AI Strategy: Powering Data Centers
Marathon's 2025 AI Power Play: Fueling Data Centers
Marathon, CPChem, Big West leverage AI to boost speed and reliability
MPLX and MARA Announce Collaboration on Integrated Power Generation and Data Center Campuses in West Texas
Opportunities for Positive ROI from the Application of AI to Refinery Processes
Marathon Petroleum Forges Robotics Partnership to Streamline Operations
Marathon Petroleum to Present DRA at AI in Energy Conference
Daniel Byrne | AI in Energy Summit
2024 Annual Report
Related Companies
About AI Tracker
AI Tracker is a research project by Larridin, the AI execution intelligence platform.
Methodology: We analyze earnings calls, press releases, partnership announcements, and product documentation. All assessments are based solely on publicly available information—no private customer data is used.
Maturity Scoring: Each dimension is rated on a 4-tier scale (Nascent → Emerging → Scaling → Leading) based on evidence from public sources. Industry averages are computed as the median across all tracked companies in the sector.