If you’ve been following application development during 2026, you may have noticed something important: AI is no longer simply helping developers write code.

It is starting to change how applications are designed, tested and delivered.

After more than 30 years working with technology and application development, I’ve seen plenty of changes. But the current shift towards AI-assisted and agentic development is certainly one of the most significant.

The important question for businesses, however, isn’t whether AI can write software.

It’s whether we are using it to build better software.

From AI Coding to AI-Assisted Development

Recent research from JetBrains found that 90% of professional developers surveyed were using AI coding agents at work at least weekly, with 68% using them daily.

That tells us something important: AI coding tools are quickly becoming part of the normal development environment.

Developers can use AI agents to generate code, investigate problems, create tests, explain unfamiliar systems and automate repetitive tasks.

That can significantly improve productivity.

But there is an important distinction between generating code quickly and engineering a successful application.

Human Expertise Still Matters

I’ve always believed that good application development starts with understanding the business problem.

AI can generate an impressive amount of code, but it doesn’t automatically understand your customers, your processes, your security requirements or your long-term business objectives.

That is why experienced developers remain essential.

Architecture, security, performance, usability and maintainability still require human judgement. Current research into agentic coding also emphasises the continuing importance of human oversight and validation.

In other words, AI should make experienced developers more capable, not make engineering discipline unnecessary.

Security Needs to Be Built In

There is another reason businesses need to think carefully about modern application development: security.

The EU Cyber Resilience Act is already introducing cybersecurity requirements for products with digital elements, with reporting obligations having started in September 2026 and the main requirements becoming applicable in December 2027.

For businesses developing software, this reinforces an approach I’ve recommended for years:

Security should be designed into an application from the beginning, not added just before launch.

What Does This Mean for Your Business?

If you’re planning a new application or modernising an existing system, don’t start with the question:

“Which AI development tool should we use?”

Start with:

• What business problem are we solving?
• Who will use the application?
• How should it integrate with our existing systems?
• What security requirements do we have?
• How will the application scale?
• How will we measure success?

Once these questions are clear, AI and modern development tools can become extremely powerful.

At Digitus Consulting, we combine modern application development techniques with decades of practical experience. Our focus isn’t simply on building applications faster. It’s on building software that is secure, scalable, maintainable and genuinely useful to your business.

If you have an application idea, an ageing system that needs modernising, or want to understand how AI could improve your software development strategy, let’s have a conversation.