Pluralsight Tech Learner Pulse: The most in demand skills so far in 2025

Want to keep up with what’s hot to learn in tech right now? We’re here to help with the Pluralsight Tech Learner Pulse, a mid-year analysis of what tech professionals are studying right now, so you can keep up to date and tailor your own upskilling from an informed place.

To make this report, we examined our database of over 2.9 million tech learners, checking how learning trends had changed since the start of 2024. We also compared these insights to predictions we made in Pluralsight’s 2025 Tech Forecast to see how on the money we were. Keep reading to find out what we discovered.

1. Cloud technologies, not AI, are top of mind for tech learners 

With all the hype around AI, you’d assume it would be the number one thing tech professionals would be upskilling in. But over the last year, the strongest interest has been in cloud technologies, and interest in this area actually spiked.

Microsoft Azure was the most popular topic amongst tech learners (+131%) while AWS was not far behind in fourth place (+89%). Google Cloud Platform had a comparatively modest uptick (+34%) and sat much further back as the 50th most popular subject among learners.

Was the forecast correct?

Yes. We predicted that cloud professionals would still be in strong demand this year as organizations struggle to launch their AI projects, manage legacy infrastructure, and deal with the sprawl of multicloud. Additionally, cloud knowledge tends to be useful in nearly every domain—such as software development, cybersecurity, AI, and data—making it an easy choice for professional upskilling.

2 Cybersecurity-related topics saw massive boom in 2025

It’s a dark time to work in the cybersecurity industry. Cyberattacks are increasing in number and sophistication, AI is upending the landscape, and there’s also geopolitical tensions and increasingly complex supply chains to deal with. Meanwhile, very few people are prepared for quantum computing, and bad actors may already be taking advantage of it.

Perhaps the biggest pain point is there’s not enough skilled people to help put out these fires. In 2025, the cyber skills gap actually increased by 8% according to the World Economic Forum, with only 14% of organizations confident they have the people and skills they need today. The result is that teams are often understaffed, burnt out, and feeling like adversaries are always one step ahead. 

So let this be a bit of good news: In 2025, cybersecurity-related topics saw massive growth across multiple areas of the discipline. Topics such as security management (+164%), identity and access management (+93%), security awareness (+164%), information security (+69%), cloud security (+93%), secure coding (+102%), and security controls (+173%).

In short, while times look grim right now, a new generation of cybersecurity professionals may be marching on the way to help.

Was the forecast correct?

Yes. We predicted there would be a strong spike in demand for cybersecurity specialists and learning these skills, spurred by advancements in AI and the increasing challenges organizations face to stay secure. 

Another potential factor in this learning spike could be professionals flocking to cybersecurity due to the perceived “future proof” nature of the profession compared to other tech fields which are experiencing AI uncertainty (E.g. Software development.)

3. Tech professionals have been furiously studying AI in 2025 (and trying to fake it in the meantime)

It’s no surprise that with 95% of organizations using AI as a hiring factor, tech professionals have started studying up in this area. Artificial Intelligence was one of the most popular subjects among tech professionals (+26%), and interest in Large Language Models (LLMs) skyrocketed in particular (+419%). In fact, searches for specific LLM applications jumped this year, such as Anthropic’s Claude (+690%), Google Gemini (+315%), and Microsoft Copilot (+43%). Curiously, there were two notable exceptions: OpenAI’s ChatGPT (-29%) and GitHub Copilot (-39%).

In Pluralsight’s latest AI Skills Report, we discovered 79% of tech workers are pretending to know more about AI than they actually do, with C-suite executives being the biggest offenders. However, given the current rate of study, it seems unlikely that this will remain the case for long.

Was the forecast correct?

Yes. We predicted the shortage of AI-specific talent would continue in 2025, leading to significant job opportunities and a rush of interest in this space. With organizations seeking to adopt AI investing more money into this space, it is natural that more learners would be taking these types of courses to deliver on their key projects and leverage these tools, particularly LLMs.

4. Ethical and responsible AI is getting love from learners as well

Last year, Pluralsight released data on how while a staggering number of people were currently learning about AI, only a fraction (1.8%) were actively searching for how to do it responsibly. 

While this hasn’t changed in 2025—nobody seems to be going out of their way to search for it unlike, say, learning Python—our data shows the number of learners actually taking ethical AI courses is actually quite high, sitting at our 35th most popular subject.

Why the difference? There are a myriad of possible explanations, such as:

• There are more dedicated AI ethics learning paths on the market as the technology matures. 

• A well-rounded AI course is likely to touch on AI ethics, meaning practitioners pick it up through unintentional exposure rather than an active choice. 

• Leaders may be actively recommending or requiring professionals to upskill in this area.

Was the forecast correct?

It’s mixed. We predicted that responsible and ethical AI adoption would be a rare and valued skill in 2025, particularly in companies in or dealing with the EU. While the value of learning about ethical AI is indisputable (Avoiding fines of EUR 35 million or 7 percent of your global annual turnover, and also avoiding public relations disasters), the fact that it is the 35th most popular subject amongst technologists indicates it is not rare but actually rather common. Whether or not this knowledge translates into practice is another matter entirely.

5. Professionals rushed to learn about Agentic AI and LangChain

In 2024, nobody was learning about Agentic AI—mostly because it didn’t really exist in any form yet. But in 2025, it has become a highly searched-for topic among tech learners, currently sitting at 59th place. 

Meanwhile, interest in LangChain—a framework for building LLM-powered applications—spiked significantly (40%). It now sits among our top 100 searched topics (currently ranking 84th.)

Was the forecast correct?

Yes. We predicted agentic AI would be used in many areas of tech from software development to infrastructure automation to business processes, with LangChain being the hot skill to learn.

6. Interest in learning containerization doubled 

The number of tech professionals learning about containerization rocketed up this year. Containerization jumped by a whopping 113%, bringing it to the 21st most studied topic. Meanwhile, Kubernetes (+26%) and Docker (+39%) also increased in learner popularity, now sitting at 19th and 32nd place respectively.

Was the forecast correct?

Yes. We predicted Kubernetes would continue to be an in-demand skill, especially as organizations used it to assist their AI/ML workloads. The need to encapsulate applications so they run in different environments with Docker is also unlikely to change in the immediate future.

7. Python increased its lead as the go-to programming language, while Java, JavaScript, and C# lagged

Just when you didn’t think Python could get any more popular as a programming language, it snaked its way even higher up the ladder. In 2025, the number of tech professionals learning Python jumped by 14%. It appears professionals have not stopped learning to program simply because they have access to AI coding assistants (at least not in Python.)

This trend on the Pluralsight platform is mirrored in the TIOBE Programming Community Index, where there has also been an 8% increase in Python popularity. That’s historic, because no programming language has ever ranked as high as Python does now on the index. Note that Pluralsight is measuring people learning Python while the TIOBE index is measuring people writing in it, indicating there’s a lot of follow-through going on. 

SQL jumped significantly in popularity by 27%, now sitting as the 9th most popular subject to learn among tech professionals. Meanwhile, other languages declined in popularity such as Java, C#, and JavaScript. 

The decline in non-Python languages was reflected in our search data, with every language but Python seeing a downturn in search queries. Interestingly, Rust—which is widely considered the most “loved” programming language amongst programmers, even if it’s not the most used— was searched by learners nearly half as much (-49%) as it was last year.

Was the forecast correct?

Yes. Honestly, this was an easy guess for us at Pluralsight; Python has absolutely dominated the programming language popularity polls for years, and AI has only made it more popular. Not only is Python easy to learn, it has a wealth of data science and machine learning libraries like PyTorch, NumPy, Pandas, and Tensorflow. 

We also predicted that SQL would continue to be in demand due to being a key skill for data professionals and its wide support by cloud providers. There were no predictions about other languages declining in 2025, but this is an interesting trend we will be keeping our eye on.

Conclusion

Even as tech professionals have dived into learning about AI, they have still not abandoned upskilling in other key areas such as cloud computing, cybersecurity, DevOps, and IT Ops. In the programming space, Python and SQL have jumped in popularity, while interest in other languages has declined in the last twelve months. 

One thing remains the same from last year, and likely for the foreseeable future: the shifting demands on tech professionals, and with it, the need for continuous upskilling to remain indispensable and current, whether it’s to meet personal aspirations or the business needs ahead.

Methodology

This analysis is based on Pluralsight platform usage data for H1 2024 and H1 2025. All percentage changes cited are year-over-year comparisons of the number of unique users either studying courses grouped by topic or entering specific search queries into the platform.