Yes, AI eats the world. But we deliberately left out a bunch of hyped topics. This radar focuses on what we believe has either (a) reached enough maturity to be used responsibly, or (b) has a realistic chance of being relevant longer than a hype cycle.
Another intentional change: we removed the “obvious / already-in-use” layer. In past editions, the many “Use” topics created comfort—but also drowned out what changed. This year, you’ll find fewer items, but each one is there because we think it actually moves decisions.
A few notes on themes in 2026.
AI in engineering: augmentation without loosing the craft
AI-augmented software development is real value now. Copilot, Claude, Junie-style tooling can augment personal skills and widen what teams can achieve. The trap is confusing speed with progress, and confidence with correctness.
Consider these tools over the whole software development lifecycle, do not limit yourself with only seeing them as code-generators. They also have capabilities over the whole quality spectrum of solutions. While they can generate throw-away scripts way faster, they also can help in supporting you finding edge cases, blind spots in parts of your code where every bit of human engineering expertise is needed and should be well thought through.
Closely related is the AI slop pattern: unverified AI output that creates downstream work for everyone else (reviewers, maintainers, ops). This isn’t theoretical – some rigorous dev productivity work shows that “AI access” can even slow experienced developers down in real tasks, despite them feeling faster. Building up expertise and experience with these tools may be even more crucial that with other tools.
A lot of current GenAI discourse optimizes for feature delivery velocity. But optimizing for speed alone builds a debt made of small fractures: duplicated logic, subtle inconsistencies, “almost correct” behaviour, and a growing review/debug tax.
Tech Trends 2026
bbv Technica Radar
Which IT technologies, tools and trends should Swiss SMEs rely on? The Technica Radar from bbv provides an answer to this question. It visualizes the IT and developer trends specifically for Switzerland.
Requirements engineering: stop paying interest in delivery chaos
Underinvesting in requirements doesn’t remove the work. It defers it with interest: unclear scope, rework, misaligned expectations, surprise dependencies.
We push a simple reframing: requirements engineering is a continuous product decision practice, done collaboratively and iteratively. Goals, constraints, success metrics, assumptions, and acceptance criteria evolving with tight feedback loops.
We do not support our observed trend in lowering the invest in requirements engineering practices. Especially when a newer hype of spec driven development which moves the requirements engineering further to the developer, you need a sound knowledge base of how requirements should be formed, formulated and discussed with the stakeholders.
Empathy & humility: the multiplier we can’t automate
Remote work, pressure, and AI hype amplify misunderstandings. Empathy is how we gather information; humility is how we let that information change our behaviour – without losing standards. Teams that keep curiosity high and judgment low will simply execute better.
This is not new, this was always the case – sometimes forgotten or left behind in the struggles of our daily life.
Let us know your additional insights and challenging thoughts.
Here you can download the whole Technica Radar Vol. 6/2026.
THIS MIGHT ALSO INTEREST YOU
Rust: Tools for reliable dependency management
Successfully introducing Rust into your team
“Not all trends make it into the bbv Technica Radar”
