How to fact-check solar safety claims in an era of industry misinformation

Not long ago, getting informed meant digging into published, often peer-reviewed material. There was friction in the process, and that friction helped filter quality. Today, information moves instantly. A short video, a social post, or a headline can spread widely without much context, and without much accountability.

That shift matters. Because now, a lot of what people see about solar safety (or any other industry for that matter) isn’t coming from structured analysis. It’s coming from content designed to get attention.

Safety narratives

Safety issues tend to get amplified for a simple reason: they’re interesting. Nobody shares a story about the millions of systems that are working exactly as designed, but one failure (one fire, one bad installation, etc.) can travel quickly.

That creates a skewed picture. The industry is full of systems quietly doing what they’re supposed to do every day, but those don’t show up in feeds or headlines. What shows up are the outliers.

Once those outliers start circulating, and they can take on a life of their own, especially when they’re repeated, reinterpreted, or framed to provoke a reaction.

Five ways to critically evaluate safety claims

Start with the sample set. One of the biggest issues in how safety claims get interpreted is where the data is coming from. If you’re analyzing failed components—connectors that burned or systems that had issues, for example—you’re naturally going to see problems or selective bias. But selective bias doesn’t tell you how common those problems are across the full population.

It’s the same dynamic you see in service. If your job is to fix broken systems, your entire view of the world is systems that are broken. That doesn’t mean most systems are failing. It just means you’re seeing a slice of the data. This slice of data can spur further inquiry and investigation to more fully distinguish symptoms from the root cause.

Look at the broader context. Numbers without scale have limited meaning. Hearing that a certain number of failures occurred can sound alarming, but the real question is: out of how many systems? Over what timeframe? Does the data illustrate a trend or data pattern that merits further investigation?

In an industry with millions of installations, a small number of incidents may not indicate a widespread issue; without that context, it’s easy to draw the wrong conclusion.

Ask who’s driving the narrative. A lot of today’s content is driven by incentives, whether that’s clicks, sponsorships, or alignment with a particular technology or approach. That doesn’t automatically make it wrong, but it does mean it’s worth asking what perspective is being represented.

In some cases, influencers sharing the loudest opinions are also financially tied to a specific outcome. That doesn’t always get disclosed clearly.

Be cautious with methodology—and lack of it. Not all analyses are created equal. Some are rigorous, with clear methods and traceable data. Others are based on limited observations or loosely defined conclusions.

If it’s not clear how the data was sourced, what was included, or what was excluded, it’s hard to treat the conclusions as representative of the industry as a whole.

Don’t confuse repetition with validation. One thing that’s become more common is seeing the same claim repeated across multiple sources. It can feel like confirmation, but often, those sources are all pointing back to the same original dataset or report. That’s not independent validation. It’s just amplification.

Real-world data

What’s changing in the industry is the scale of data we now have access to. With connected systems and monitoring platforms, there’s a much clearer view into how systems are actually performing over time.

That kind of large-scale, real-world data is critical. It helps move the conversation away from isolated examples and toward building understanding, and patterns across entire fleets. Instead of asking, “Did this fail?” the question becomes, “How often does this fail, and under what conditions?”

AI: Helpful, but not a source of truth

AI is starting to play a bigger role in how people gather information. AI can be a powerful tool, but with it comes new challenges. These systems are trained on large datasets that include everything from high-quality research to social media posts and forum discussions.

The result is that AI can produce answers that sound confident and well-structured but aren’t always grounded in verified data. It doesn’t tell you how confident it is, or where the information really came from. In that sense, it’s a useful tool, but not something to rely on without doing deeper validation.

A grounded way forward

The solar industry isn’t short on data. If anything, it’s the opposite. The challenge now is making sense of it. That means taking a step back from initial reactions, asking where information is coming from, and looking at the bigger picture. It means recognizing that outliers exist but also understanding how they compare to the broader system.

Because at scale, safety isn’t defined by a handful of incidents. It’s defined by how millions of systems perform, day in and day out, and understanding that difference is what separates noise from signal.

About the author: Noah Tuthill is the Director of Codes and Compliance at SolarEdge, with more than 20 years of experience across the solar and renewable energy space. Based in New England, he works on North American codes and standards, product certification, and collaboration with Installers, utilities and AHJs, drawing on experience with UL and IEEE standards, NEC requirements, and grid interconnection. His career started in the field and has since spanned engineering, R&D, and product development, which helps keep his work practical and grounded in real‑world experience. Outside of SolarEdge, Noah enjoys travel, adventure, and spending time outdoors—and especially likes sharing those experiences with his wife and their twin boys.

The views and opinions expressed in this article are the author’s own, and do not necessarily reflect those held by pv magazine.

This content is protected by copyright and may not be reused. If you want to cooperate with us and would like to reuse some of our content, please contact: editors@pv-magazine.com.