Offshore Europe 2025: DNV outlines containment measures for threat posed by AI

DNV will deliver a presentation at the SPE Offshore Europe 2025 conference on the risks arising from unpredictable interactions between humans and artificial intelligence (AI) tools, the NOF organization reported Sept. 1.

The paper, to be presented on Sept. 3 as part of the conference’s "Digital: AI Supported Subsurface" technical program, will highlight DNV’s research into the risks that AI poses in the offshore sector. 

As the paper notes, AI technology itself continues to evolve rapidly, with ever-increasing capabilities and potential applications.

“Ensuring safety in the face of a disruptive new technology is paramount, particularly in the offshore environment. AI will certainly find applications within the sector; however, its emergence presents a real challenge to regulators due to its disruptive nature, extremely rapid pace of development and easy accessibility to individuals. For example, if an individual engineer uses a generative AI engine to write standard operating procedures for a new piece of equipment, how does the industry ensure that maximum safety is maintained? This is just one example of how AI-related risks could potentially overwhelm traditional safety mechanisms.”

—Graham Faiz, Head of Digital Energy, DNV

To avert the possibility of AI-related risks overwhelming traditional safety mechanisms, DNV is proposing a systemic framework that incorporates safety engineering principles and AI assurance throughout the full system lifecycle. 

It plans to test AI in offshore settings to help the industry understand and manage safety risks, with use cases devised and studied to help identify and suppress dangers. The focus will be on how AI changes over time, determining the complexity of its decisions, and how it affects both personnel and the wider system. 

DNV will use this approach to form the basis of the safety framework.

“AI is becoming increasingly ingrained in the industry, and as humans and AI become more integrated their interactions become more unpredictable. However, we have seen little consideration of how existing safety systems can address this new technology. A new approach is needed to directly address the emergent risks of AI integration across our industry. ... By identifying and mitigating potential risks, our work aims to prevent unintended consequences and ensure the safe integration of AI in safety-critical operations.” 

—Hari Vamadevan, SVP and regional director UK and Ireland, Energy Systems, DNV

A DNV survey found that, of 1,160 companies canvased across Europe, North America, Central and South America, and Asia, 60% of respondents affirmed that they would use AI-driven applications in their operations going forward, up from 47% in 2024.

DNV invests in R&D facility

DNV also revealed that it has invested in upgrades of its Spadeadam R&D center in Cumbria, Northwest England, to support de-risking of the energy transition.

One aim is to help demonstrate the safety of technologies for both UK-based and international organizations aiming ultimately for net-zero emissions from their operations, DNV reported Sept. 2.

The Spadeadam facility provides immersive technology, incuding 4k streaming that allows people worldwide to watching demonstrations happen in real time.

For over 45 years, the site has been used to perform research into the potential consequences of hazardous incidents and their impact.

The new investment will allow operators, regulators and technology providers to jointly develop safer renewable energy systems by combining conventional oil and gas expertise with emerging technologies such as hydrogen and carbon capture and storage.

In addition, the site is hosting DNV’s Skylark joint industry project, which seeks to improve safety standards in CO₂ pipeline operations.

DNV is exhibiting at booth 2L60 at the SPE Offshore Europe event this week in Aberdeen, Scotland.