The Future of EMC in the Age of AI
Written by: Anton Tishchenko
January 10, 2026
Welcome to 2026! The new year has caught me in the midst of my planned career break, which followed the timely submission of my PhD thesis. So, I thought this may be a good opportunity for me to practice blogging and writing about my decade-long experiences in the EMC, RF, and other aspects of the electronics industry. I would like to open with a not-so-technical, yet highly controversial topic that I’m sure concerns every electronics engineer on the planet in 2026 – that is, the use of AI tools in engineering and the prospective redundancy of human skills, concerning PCB design, circuit design, and EMC/RF simulations. The goal of this entry is to provide an outlook for the year and list some of the key emerging technologies that you should watch out for.
Firstly, let’s take a look at the definition of AI. In the Oxford English Dictionary, AI is defined as “the capacity of computers or other machines to exhibit or simulate intelligent behaviour.” There’s an interesting word in this definition – “intelligent”. This word is subjective, as some of us may interpret something as intelligent, while others may not. As per the Oxford English Dictionary, the earliest known use of the word intelligent was recorded in the early 1500s, and the most recent definition is given as: “having a high degree or good measure of understanding; quick to understand; knowing, sagacious.” This definition gives us some metrics that we can use to apply our subjective judgment and decide whether the AI-branded tools that we encounter are truly becoming more “understanding; knowing, and sagacious (able to provide good judgement)” or not.
Let’s consider a quick example. If I had asked my premium version of ChatGPT to write up a 2-page article about “The Future of EMC in the Age of AI”, listing the most recent trends and providing an outlook for the year 2026, would it be capable of knowing and understanding my requirements and then sagaciously producing a worthy blog entry? I don’t think so. Would it ever be capable of performing such a task in the near or distant future? Having seen the implementations of AI hardware under the hood, I highly doubt that as well. After all, I am writing about my personal experiences, fueled by a 10+ year- long career in engineering, and unless all this knowledge can be extracted from my head into a sufficiently large dataset, I simply do not see a way for a reasoning AI to exist, beyond the hype of marketing tools. If such a possibility were to really exist, then who’s to say that the resulting intelligence is artificial? I’m getting all philosophical now. But let’s talk about the simple AI tool that I’m using as a part of my writing process. I’m talking about Grammarly, of course. I installed this simple tool several years ago and never looked back. What does Grammarly do? It checks my English grammar and offers suggestions for improvement. How does this tool know and understand what suggestions to make? Well, its decision-making process is backed by the rules of the English language, which were set out by the experts and developed over centuries of continuous use. Yet, it is also changing, as the new words and phrases appear, which often require new rules as well. So, this ability to learn and to perform a case-by-case analysis of the English language – this is what makes Grammarly truly intelligent, and different from a simple calculator-like logic.
We can now draw parallels with electronic engineering and its own use cases for AI. The real value of AI in electronic engineering is to enforce the same disciplined process experienced engineers already use (i.e. best design practices), then validating it automatically and suggesting improvements. This formula applies to many aspects of electronic engineering, including circuit design, PCB design, Design for Manufacture (DFM) checks, and EMC analysis. The use of AI here, for example, can be:
● To convert a circuit schematic from the application datasheet into a usable block that can be directly imported into the CAD (check out Celius and Circuit Mind). However, remember about GIGO - garbage-in produces garbage-out, so selecting the training source carefully is hugely important when using these tools. Many application datasheets are outdated and give wrong advice, which is unknown to AI.
● Another example can be the use of automated PCB routing and component placement (check out JITX, Flux and DeepPCB ), yet remember that by design, AI is compelled to produce an answer even if the answer is wrong. This often results in unrealistic designs, caused by the lack of constraints, requiring many hours to be spent setting up the design properly, which could easily be spent on doing the layout yourself.
● Finally, the most promising application of AI in electronics engineering, in my opinion, is the use of AI for DFM checks and enforcing other constraints, just like how Grammarly checks the English grammar in the previous example. This application lends itself naturally towards automated checks for potential EMC issues (check out Denpaflux). Soon, PCB for EMC will provide its own competitive solution in this area, which is currently available for beta testing only. Stay tuned.
In this blog entry, I’m not going to make any comparisons between these solutions. However, the impact on the electronics industry is already noticeable. If you search for jobs on LinkedIn with the keyword “Junior Electronic Engineer”, do not expect to get many hits. The use of AI tools has wiped out the majority of entry-level jobs in the market, and you can expect this trend to continue in 2026. Applicants are expected to have at least 5 years of experience, which casts doubt on the value of higher education and degrees in electronic engineering. Internships, apprenticeships, and diplomas often provide a much faster route to the job market, especially given that it has never been easier to learn circuit design than now. This may be a controversial statement, but I believe that electronic engineers are likely to become technicians in the very near future, with companies developing their own AI tools for design assistance to guide these technicians. Only the AI developer-level jobs would require the level of education, typically provided by academic institutions. Having said that, the human aspect of electronic engineering cannot be phased out. Just like every blog post that is worth reading is unique, every PCB and every product that is worth manufacturing is also unique. The use of AI tools will never substitute product design entirely, but the way that it is done in 2026 might be quite different from what we have become used to. The educated use of AI tools can drastically reduce the number of human errors in the design, cutting down the average number of required revisions from 3-5 to 1-2, and thus speeding up the release to market. The overall quality of consumer electronics is expected to increase with the widespread use of AI, as the bad design practices, such as split ground planes and star grounding, will sink to the bottom of the ocean of human knowledge. This will also foster the development of rapid prototypes, increase the number of first-time EMC passes, and reduce the likelihood of companies getting stuck in the R&D hell cycle, making what was previously thought of as complex and requiring specialist knowledge easy and accessible to anyone.
In summary, I believe that 2026 will be a good year for electronic engineers and the industry. However, staying competitive in the age of AI will become a challenge that many of us will have to face. Naturally, the centres of engineering activities are likely to move to the emerging economies such as the Philippines, Indonesia, India and, to a lesser extent, China. To hire someone to do a relatively simple PCB layout job in a high-income country may become a luxury that only the Defence and National Security industries will be able to afford. In this environment, electronic engineers may have to step up their game and pick up new skills, focusing on production management and software development.
About the author
Anton Tishchenko received the B.Sc. degree in sound engineering from Wrexham University in 2015, and the M.Sc.(Distinction with Hons.) degree in electronic engineering from the University of Surrey in 2021, where he is currently pursuing a PhD degree within the 5G and 6G Innovation Centres, Institute for Communication Systems, focusing on reconfigurable metamaterials and metasurfaces for the next generation of wireless communication networks. Before his academic work, he established himself as an electronic/RF design specialist, developing hardware for the professional audio, telecommunications, and intelligent transportation industries. As a part of his Principal Electronic Design Engineer role at Cubic Transportation Systems Ltd, he has led several multi-disciplinary engineering teams throughout his career, on Transport for London (TfL), Network Rail, and MTA projects. He hosts an RF-dedicated YouTube channel, “Dr EMC”.
- January 11, 2026
- 10:20 pm
Some hard facts were stated here, but in action I see the opposite! A lot of commercial companies realized how dangerous it is to outsource work to NATO Hostile countries, and they are now insourcing.