Offshore energy industry now utilizing AI for critical upstream operations

Glenn Legge, Endeavor Management 

The offshore energy sector is now utilizing artificial intelligence (AI) for critical upstream operations.

Recent examples include:

• Schlumberger’s (now SLB) use of AI for deepwater hydrocarbon reservoir characterization in association with the Abu-Dhabi National Oil Company (ADNOC)
• Schlumberger’s AI-assisted DrillOps automation and DrillOps advisory tools.
• Halliburton’s AI-powered LOGIX platform to enhance drilling performance and reservoir recovery.
• Woodside Energy awarded a contract to Schlumberger to utilize AI-enabled drilling technology for the deepwater Trion project offshore Mexico.

But while AI utilization can provide substantial performance and cost benefits, there are risks. The risk exposures from AI usage are still being determined as AI capabilities continue to rapidly evolve. Companies utilizing AI-assisted services must develop the capabilities to analyze, anticipate and manage risks that can arise from the use of Gen AI and AAI technologies.

But before identifying risk, the offshore energy industry must address the different capabilities and utilities of Gen AI and AAI:

• AI is making subsurface imaging of hydrocarbon resources in the Gulf of America faster and more precise.
• AAI can also utilize user-provided information, as well as conduct proactive, independent research and analysis to create substantive plans/strategies to execute projects or goals.
• Predictive maintenance of critical exploration and production equipment through the use of AI-assisted digital twin technology.

AI is also being utilized by the offshore energy sector, and its regulators, in assessing the integrity of offshore energy structures in the Gulf of America, some of which are overdue for decommissioning. These structures may be repurposed for alternative energy projects, including:

• CO₂ transport, injection and storage.
• Locating/constructing and operating offshore data centers in the safest and most cost-effective manner.
• Alternative energy operations involving the production of Green and Blue H₂.

Still another example came recently from the US Department of Energy (DOE). In October 2025, the DOE issued its Artificial Intelligence Strategy update in October 2025, acknowledging the use of the National Energy Technical Laboratory (NETL) Advanced Infrastructure Integrity Model (AIIM) to evaluate energy infrastructure integrity for maintenance and planning forecasting. Per the DOE  update, the AI program can be used on natural gas and other pipelines, as well as offshore platforms and wellbore integrity.

Internal risk management

This author’s previous article in the September-October issue of Offshore provided an overview of managing external AI exposures through contractual risk and insurance coverages to address third party claims. In this article, we consider:

• Effectively assessing and managing internal corporate AI risk exposures, and
• Whether companies are realizing the intangible asset value associated with the beneficial utilization Gen AI and AAI.

Internal operational risk assessment and management of Gen AI and AAI utilization differ substantially from most “external” exposures that can be readily anticipated and addressed through contractual risk allocation strategies and appropriate insurance coverages. The rapidly evolving capabilities of Gen AI and AAI are more challenging to classify and manage than the more traditional operational exposures in the offshore energy sector, such as integrity of critical well control equipment and determination of reservoir integrity.

Initially, risk management programs for AI often consider determining:

• The analytical capabilities of the AI platform that is being utilized
• A functional framework/scope of work for AI utilization
• The scope of data that AI mechanism will ingest and utilize to create its work product
• The reliability of AI generated analyses/work product.

As set forth above, Gen AI is “fed” data by the user, therefore companies utilizing Gen AI can determine the scope of data that will be involved in the AI analyses.  AAI, on the other hand, has the capabilities to “hunt and gather” information autonomously, which may or may not be “pre-checked” by the user. In addition, the use of Gen AI and AAI may create unanticipated risks for users due to occasional “hallucinations” that can adversely impact the outcome/strategies generated by these processes. 

The offshore energy sector has become familiar with the operational parameters for certain Gen AI utilizations, and has addressed basic risk management criteria, such as:

• Establishing an operational framework for AI governance that addresses appropriate data utilization and the AI responses/strategies based upon this data.
• Controlling access to relevant, validated, and appropriate data and preventing inappropriate access to non-relevant or inappropriate data such as:

• personal identifiable information (PII) of employees or third parties
• proprietary/confidential entity information, such as intellectual property, business strategies or private financial analyses.

• Utilizing risk management strategies and capabilities that can evolve rapidly and continuously to address a variety of exposures related to AI data consumption and risks of hallucinations.
• Managing internal corporate access to Gen AI and AAI usage.

Internal AI risk management programs must have the capabilities to continuously monitor the development of Gen AI and AAI capabilities/exposures and the ability to effectively address any unanticipated risks.

External risk management

The significant capabilities of Gen AI and AAI, and the external exposures that can arise from those applications must continually be addressed. Examples of these exposures could include:

• Use of AI in critical, high-pressure deepwater drilling operations that could result in significant natural disasters and regulatory impacts.
• AI utilization in creating/managing large engineering, procurement and construction (EPC) contracts that can have valuations and exposures in the  billions of USD.
• Utilization of AI autonomous inspection and predictive maintenance of critical equipment.

To make the strategic evolution of AI utilization even more complex, companies utilizing Gen AI or AAI may be subject to cyber-attacks that are also utilizing generative AI platforms such as WormGPT. These AI-assisted cyberattacks can utilize polymorphic malware that can evolve in an effort to evade existing and updated software and cybersecurity programs. The significant cost of implementing cybersecurity that can protect Gen AI or AAI programs from rapidly  evolving cyberattacks must also be considered in realistic ROI projections.

The scope and cost of cybersecurity for Gen AI and AAI platforms will be impacted by the guidelines and requirements of both the US National Security Agency’s Artificial Intelligence Security Center (AISC), the US Cybersecurity and Infrastructure Security Agency (CISA) and the Federal Bureau of Investigation (FBI).

In May 2025, CISA issued Guidelines for Secure AI Systems Development, and a Cybersecurity Information Sheet, in association with the FBI and various foreign government cybersecurity agencies. The documents were focused on “ensuring the accuracy and integrity of AI outcomes an outlines potential risks arising from data integrity issues in various stages of AI development and deployment.”

Creating tangible value

Recent industry surveys from Gartner indicate that the most significant hurdles to companies utilizing AI in core activities are “high costs, immature technologies and organizational resistance.” A June 2025 Gartner report projected that 40%, or more, agentic AI projects may be canceled due to lack of strategic implementation, costs and unclear projection of ROI. This data indicates that a company’s efforts to successfully manage AI is a challenging goal – one that may not be achieved by the majority of companies utilizing AI.

Finally, we should address whether a company’s effective management of AI processes and AI risk management policies, are “intangible assets” that create substantive value for the business entity.

Recently, Endeavor Management met with officials from Andersen Consulting to discuss the conceptually challenging relationship between intangible assets and enterprise value. 

During the discussion, it was observed that companies provide data, which is likely an intangible asset, to AI; or they define the scope of data that AAI will obtain. Gen AI and AAI will then generate beneficial data, resources and/or strategies that may benefit the company’s brand value, reputation, business relationships and regulatory compliance records. These are all benefits that should create tangible value for a corporate business entity.

The company should then determine the tangible value of this beneficial use of these processes and intangible assets, then move toward successfully managing AI for the benefit of the company and its customers.