In 2025, AI will reshape the way teams think, build, and deliver software. We’re now at the point where “AI coding assistants have rapidly moved from novelty to necessity (up to 90% of software engineers provide us with some kind of AI for coding),” says Eddy Osmani. writesIt’s a very different world to the one we were in 12 months ago, As we look toward 2026, here are three key trends we see driving change and how we think developers and architects can prepare for what’s next,
Developing coding workflow
New AI tools change the coding workflow in 2025, enabling developers to write and work on code faster than ever before. This does not mean that AI is replacing developers. It’s opening up new frontiers to discover and skills to master, something we discovered for the first time ai codecon In May.
AI tools have been revived in IDEs and on the command line. Discussion Regarding the future of IDEs, echoing previous arguments (e.g., VS Code vs. Vim). It is more useful to focus on the purpose of the tool. As discussed by Kent Beck and Tim O’Reilly in NovemberDevelopers are ultimately responsible for the code produced by their chosen AI tool. We know that LLM “actively reward Current top level software engineering practices” and “enhance existing expertise“As told by Simon Willison. And a good coder”Include questions AI doesn’tDoes it really matter which device is used?
The key transferable skill to working with any of these tools is to understand how to effectively communicate with the underlying models. AI tools produce better code if they are given all the relevant background on a project. Managing what AI knows about your project (reference engineering) and communicating it (speedy engineering) are going to be the key to doing good work.
The core skill of working effectively with code will not change in the face of AI. Understanding code review, design patterns, debugging, testing, and documentation and applying them to the work you do with AI tools will make the difference.
The rise of agentic AI
With the rise of agents and Model Context Protocol (MCP) in the second half of 2025, developers gained the ability to use AI not just as a pair of programmers but as an entire team of developers. speaker here Coding for an agentic world New discoveries unveiled at live AI CodeCon event in September 2025 tool, workflowsAnd mercenaries Who are shaping this emerging discipline of agentic AI.
Software engineers are not just working with single coding agents. They are building and deploying their own custom agents, often involving complex setups. multi-agent scenario, Teams of Coding AgentsAnd agent swarmThis change is due to the operation of AI Orchestrating AI The importance of actually understanding how good software is built and maintained increases.
We know that AI generates better code with contextand this is also truth of agentsLike coding workflow, this means understanding the engineering context is essential, However, the differentiator for senior engineers in 2026 will be how well they apply intermediate skills such as product thinking, advanced testing, systems design and architecture in their work with agentive systems,
AI and Software Architecture
We started 2025 with our January Superstream, Software Architecture in the Age of AIWhile speaker Rebecca Parsons explored the architectural implications of AI, she bluntly stated that “given the pace of change, it may be out of date by Friday.” By the time of our superstream augustThings got a little more robust and our speakers were able to share AI-based patterns and antipatterns and explain how they tie in with software architecture. Our December 9 event will look at enterprise architecture and how architects can navigate the impact of AI on systems, processes, and governance. (Registration is still open-save your seat.) As these events show, AI has moved beyond being some architect It is possible to consider what is now Necessary For their work.
We are seeing successful AI-augmented architectures being used event-driven modelAI enables agents to act on incoming triggers rather than fixed signals. This means that it is more important than ever to understand event-driven architecture concepts and trade-offs. In 2026, topics aligned with evolving architectures (evolutionary architectures, fitness functions) will also become more important as architects look for ways to modernize existing systems for AI without derailing them. AI-native architecture The next year will also bring new ideas and patterns for systems design, as will the trend toward agentic AI.
As was the case with their engineer colleagues, architects still have to know the basics: when to add an agent or microservice, how to consider cost, how to define boundaries, and how to act on the knowledge they already have. As noted by Thomas Bates, Sarah Wells, Aaron Stiller and Daniel Bryant informationQ“They too need to understand How does an AI element relate to other parts of their system: What are the inputs and outputs? How can they measure performance, scalability, cost, and other cross-functional requirements?”
Companies will continue to decentralize responsibilities across functions this year, and AI will bring new sets of trade-offs to consider. It is true that regulated industries are understandably careful about providing access to their systems. They are implementing AI more carefully, with more security and governance, but they Are It is still being released. So there’s never been a better time to understand the foundations of software architecture. This will prepare you for the complexity that may be on the horizon.
Strong foundation matters
AI has changed the way software is created, but it hasn’t changed what makes good software. As we enter 2026, the most important developer and architecture skills won’t be defined by the tools you know. They will be defined by how effectively you implement decisions, communicate intent, and handle complexity when working with (and sometimes against) intelligent assistants and agents. AI rewards strong engineering; It does not replace it. This is an exciting time to be involved.
Join us at the Software Architecture Superstream on December 9 to learn how to better navigate the impact of AI on systems, processes, and governance. Over four hours, host Neil Ford and our lineup of experts including Metro Bank’s Anjali Jain and Philip O’Shaughnessy, Versal’s Dom Sipowicz, Intel’s Brian Rogers, Microsoft’s Ron Abelera, and Equal Experts’ Lewis Crawford will share their hard-earned insights about building adaptive, AI-ready architectures that support continuous innovation, ensure governance and security, and seamlessly align with business goals. Are aligned with.
O’Reilly members can register herenot a member? Sign up for a 10-day free trial Before you join the program—and explore all the other resources at O’Reilly.
