Industrial AI, natural catastrophes, energy and supply chain resilience: 8 things to watch out for in 2025
2024 will be remembered with mixed feelings, according to Hannah Witzel, Operations Manager at FM in Germany.
“In addition to the great sporting events of the European Championships and the Olympics, there was severe destruction caused by natural disasters,” Witzel says. “This reminds us that the energy transition must continue to be a priority in the new year 2025, even if the German economy is in a tense situation and new elections are imminent.”
Witzel continues: “Another major development of our time is artificial intelligence. Companies need to ask themselves how they deal with AI - they should take advantage of the opportunities, but also recognize the potential risks. With this in mind, the experts at FM have given it some thought and formulated six predictions and insights on what’s to come in 2025 – on energy, climate, AI and more.”
1. AI AND AUTOMATION PENETRATES INDUSTRY
– LOU GRITZO, Ph.D., chief science officerLook for AI to spread from the office to growing empowerment of automation in power plants and factories, creating new property risks even as it improves labor shortcomings, manufacturing productivity, quality, and agility. AI tools will also compound an already voracious demand for large data centers and additional power generation and transmission.
Also expect:
- Louder calls to ramp up energy generation and infrastructure, including small modular nuclear reactors, potentially bringing ample electricity closer to the point of use, but posing new risks.
- Geopolitical upheaval to affect supply chains, chip supply, and raw materials for batteries and potential nuclear fuels.
- Relaxation of regulation across chemical processing, nuclear energy, and AI. A general trend of being responsive to adverse AI consequences rather than proactive policy.
- Continued focus on adapting to severe weather, such as flooding and more severe wildfires.
2. AI CYBER CONFLICT
– Roland Schaefer, Ph.D., vice president, research director of automation and cyberIn 2025, generative AI is likely to spawn more cyberattacks of more varieties on potentially larger scales than ever given its ability to automate attempted breaches. Industrial controls will remain an attractive attack surface, having older, weaker, sometimes nonexistent security protections.
Fortunately, AI will simultaneously improve cyber defense. In fact, researchers will create digital models of industrial networks to pit white-hat and black-hat hackers against each other to improve leaders’ understanding of vulnerabilities.
“As cyberattacks get more sophisticated, expect more attacks with physical consequences, as we saw in the Ukraine blackouts,” says Schaefer. “Also expect business to focus more on protection through compliance with the NIS-2 directive and better visibility on attacks on industrial controls. Until now, these intrusions haven’t been as thoroughly catalogued as traditional enterprise network attacks.”
3. RENEWABLES DEMAND RESILIENCE
– Doug Patterson, FM senior vice president, forest products and renewable energy; and Michael Perron, renewable energy market lead, FM Renewable EnergySunlight and wind are free, but renewable energy production today is vulnerable to disruption, including natural hazards and fire. In 2025, the renewable energy industry will work aggressively to protect renewable facilities and infrastructure by addressing risks such as:
- Hail, wildfire and windstorms (that can damage solar installations).
- Lightning and structural collapse (that can damage turbines).
- And thermal runaway (that can ignite battery energy storage systems) leading to super-hot, dayslong fires that resist normal suppression.
“In the coming year, look for improvements in renewable-energy research, testing (like our hail cannon), engineering, product quality, construction and standards development,” says Patterson. “As always at FM, we’ll use science and engineering to address renewable-energy risks just as we’ve addressed hurricane threats to buildings.”
4. AI GENERATES MORE VALUE FOR THE INSURANCE DOLLAR
– Pentti Tofte, senior vice president, innovation, analytics and AIIn 2025, artificial intelligence will help FM clients maximize policyholder value by better understanding risk, loss, resilience and loss prevention.
The key to helping clients extract value is by using large quantities of relevant data that AI can learn from. At FM, for example, we have captured more than 1 billion data points over 15 years detailing:
- The properties of our multinational clients and the ways they’re built, engineered and protected from loss.
- The specific losses these properties are likely to experience due to natural hazards, climate change, fires and/or equipment failures.
This data fuels FM’s proprietary AI models. “We’re getting to a point where we can predict with surprising accuracy which losses are most likely to strike at which client locations,” says Tofte. “That takes guesswork out of clients’ capital-investment decisions and dramatically reduces the chance of a catastrophic business disruption. It also gives clients more value for their premiums.”
5. CLIMATE RISK CLARIFIED
– Angelika Werner, Ph.D., research director of climate risk C resilienceAs we head into 2025, climate change remains a complicated topic, and businesspeople sometimes think that only scientists can grasp the mechanics, risks and likely effects of a warming planet. FM is committed to helping untangle the causes and effects of changing climate risk, providing actionable insights and solutions so clients can prepare before a catastrophe hits.
That’s never been more important, as recent events showed: 2024 is expected to be the warmest year on record, exceeding 1.5°C above pre-industrial average for the first time. It was also a year of floods. In September and October alone, record rainfalls hit Southeast Asia from Typhoon Yagi, Southern Spain from a DANA system and the southeastern United States from Hurricane Helene.
As we move into 2025, the general conditions for high activity persist. Global ocean temperatures remain unprecedentedly high for the second consecutive year, providing fuel for convective systems like monsoons, tropical cyclones and cut-off lows.
FM climate researchers will continue looking into trends of climate extremes activity leading to changes in rain, hail, wildfires or wind patterns globally. In 2025, FM will also continue to make significant investments in improving flood capabilities, like modeling riverine overtopping or detailed location assessment of flood impact. For severe convective storms, a globally consistent approach to hail risk is on the agenda for the year ahead, as is forming an understanding of peril- specific effects of thunderstorms such as lightning, tornadoes or heavy rain.
“While understanding your climate risk and how it is changing over time is essential, the best preparation to minimize impact is to know how to act in the light of an approaching event,” Werner says. ”FM continues to work on solutions that help avoid losses and guide the implementation of early emergency responses onsite.”
6. DATA-DRIVEN INSIGHTS
Do you have questions about protecting your business to ensure your prosperity in the future? Like Which countries are most vulnerable from a supply chain standpoint and how is climate change impacting the countries where I conduct business? In 2025, the FM Resilience Index and enhanced capabilities in understanding supply chains will seek to answer those questions.
Insurance partners like FM will also use generative AI and other AI tools to improve the overall value to policyholders through enhanced policy review processes, leading to greater contract certainty and the ability to more precisely underwrite clients’ risks and allocate insurance capacity.
“AI factors heavily in our 2025 outlook,” says Tofte. “We are cognizant of the risks as well as the opportunities. AI is enhancing our ability to protect our clients from catastrophic loss. At the same time, AI brings new emerging risks and considerations around ethics and governance, so we have a significant focus on responsible AI practices.”