Push for safe incorporation of AI in energy sector

New initiatives are being launched to enhance the safe and trustworthy development of artificial intelligence (AI) for the energy sector and more broadly.

Among these is a new Executive Order from US President Joe Biden on the development and use of AI and a suite of recommended best practices from assurance and risk management specialist DNV.

In broad terms the Executive Order, which is focussed on the ‘safe, secure, and trustworthy development and use of artificial intelligence’, is stated to establish new standards for AI safety and security and promote innovation and competition while including protections for individuals and data.

In a subsequent statement the US Department of Energy committed itself to playing a crucial role in the carrying out of the Executive Order.

Have you read?
What’s on the technology radar for Europe’s DSOs?
‘We must demystify new tech like artificial intelligence’: Viviana Vitto

Among these, the DOE will develop tools to evaluate AI capabilities to generate outputs that may represent nuclear, nonproliferation, biological, chemical, critical infrastructure and energy-security threats or hazards as well as model guardrails that reduce those risks.

The DOE also commits to collaborating with its computing capabilities and AI testbeds to build foundation models that support new applications in science and energy and for national security.

Further, the DOE will work to establish a pilot to enhance training programmes for scientists, with the goal of training 500 new researchers by 2025.

“There are certain things that we know AI is going to be great for,” US Secretary of Energy Jennifer M. Granholm was quoted in a statement.

“And, hopefully, a lot of machine learning will help to make work better for workers. But we want to make sure even as we bring all these supply chains home and all of these factories that we’re thoughtful about how AI is going to affect the life of workers across the country.”

AI recommended practices for the energy sector

DNV has set out a set of nine new or updated recommended practices that cover the four building blocks of industrial AI systems, i.e. data, sensors, algorithms and digital twins.

The advent of AI requires a new approach to risk, the company states.

Whereas conventional mechanical or electric systems degrade over years, AI-enabled systems change within milliseconds. Consequently, a conventional certificate provided by DNV, which normally has a three- to five-year validity, could be invalidated with each collected data point.

This necessitates a different assurance methodology and a thorough understanding of the intricate interplay between system and AI, allowing for a proper assessment of failure modes as well as potential for real-world performance enhancement.

DNV also highlights the forthcoming EU Artificial Intelligence Act as a first, and states that the recommended practices are intended to provide a practical interpretation of the Act.