The Shifting Sands of AI in Software Engineering: A 2023 vs. 2024 Gartner Hype Cycle Analysis

Generative AI in software engineering is evolving at breakneck speed, and Gartner ’s Hype Cycle for 2023 and 2024 show just how fast the landscape is shifting. What was once riding high in the Peak of Inflated Expectations either surged forward, plateaued, or, in some cases, disappeared entirely. Meanwhile, new entrants made their way into the 2024 cycle, signaling fresh directions for AI in software engineering.

Let us break it down: what moved across, what fell off, what was new, and where are we headed in 2025 and beyond.

2023 vs. 2024: Trends on the Move

Some AI-powered trends in software engineering have progressed across the Hype Cycle, indicating increased adoption, maturing technologies, and shifting expectations:

• AI-Augmented Software Engineering moved forward, signifying that organizations are getting past the hype and into real-world adoption. More AI-assisted tools, such as GitHub Copilot, Tabnine, Cursor and Codeium, are integrating into DevOps pipelines, supporting code refactoring, bug detection, and automated recommendations.
• Cloud-Native AI also progressed, as businesses continue integrating AI into cloud-based development workflows. The rise of AI-powered cloud platforms like AWS CodeWhisperer, Google Vertex AI, and Microsoft Azure AI suggests a shift toward scalable, on-demand AI capabilities embedded into modern software architectures.
• Code Generation and AI Coding Assistants remain critical but are now facing increasing scrutiny around security, compliance, and maintainability. The initial fascination with tools like GitHub Copilot is evolving into deeper discussions around intellectual property risks, bias in training data, and productivity trade-offs. Regulatory concerns are prompting companies to adopt AI governance platforms like Credo AI and IBM's AI Ethics Framework.

These shifts suggest that companies are no longer just experimenting—they are integrating AI into their software development lifecycles in more structured ways, looking for efficiency and sustainability rather than just novelty.

Trends That Vanished: What Dropped Off in 2024?

Some technologies that appeared in the 2023 Hype Cycle have disappeared in 2024. This could mean one of two things: they’ve either matured into mainstream adoption or fallen out of favor.

• AI-Augmented Software Testing appears to have blended into broader AI-driven development approaches, rather than being tracked separately. This suggests that testing capabilities are increasingly seen as an inherent part of AI-powered software engineering rather than a standalone innovation.
• Some early-stage experimental tools likely consolidated into broader AI-augmented software engineering platforms, especially as companies seek end-to-end AI solutions rather than fragmented toolkits. Tools like Diffblue (AI-powered unit test generation) and Mabl (self-healing test automation) are now more seamlessly integrated into larger AI-powered DevOps ecosystems.
• Technologies that failed to deliver practical ROI may have been quietly sidelined as organizations prioritize real-world impact over theoretical potential. Some AI-powered low-code/no-code platforms, for instance, while still evolving, are struggling with scalability and adoption beyond non-technical users.

The New Players in 2024: Why They Matter

The 2024 Hype Cycle introduced some new trends that signal fresh directions for AI in software engineering:

• More emphasis on Responsible AI and governance—AI in coding is no longer just about speed; it needs to be about trust, explainability, and compliance. Companies are investing more in AI observability, model interpretability, and ensuring AI-generated code adheres to ethical and security standards. Platforms like Fiddler AI, Arthur AI, and #ExplainableAI frameworks from Google and Microsoft are becoming essential.
• Deeper integrations between AI and DevOps—automation is extending beyond code generation into testing, deployment, and monitoring. AI-driven CI/CD pipelines, including AI-powered testing tools are becoming more common, enabling continuous improvement with predictive analytics and self-healing mechanisms.
• AI-powered software design and architecture—beyond just coding, AI is playing a bigger role in software architecture decisions, suggesting an evolution toward AI-enhanced system design frameworks. Enterprise software solutions are incorporating AI-driven decision-making to improve performance, scalability, and security. Tools like IBM Watsonx Code Assistant and Google’s AI-powered system design tools are leading the charge.

These additions highlight how AI in software engineering is shifting from flashy, headline-grabbing capabilities to sustainable, enterprise-grade solutions that prioritize robustness, security, and usability.

2025 and Beyond: What Next for AI in Software Engineering?

Looking ahead, expect:

• More AI-driven automation in software engineering, covering everything from architecture design to self-healing code. AI will not just generate code but will also optimize performance, suggest design patterns, and refactor legacy applications autonomously. Platforms like DeepCode and CodeQL are already making headway in AI-driven code optimization.
• A shift from AI coding assistants to AI co-developers, with more advanced capabilities in debugging, refactoring, and optimization. AI will move beyond simply suggesting code snippets to understanding high-level business requirements and assisting in system-wide optimizations. Expect tools like OpenAI’s GPT-powered agents and Meta’s AI research to drive this shift.
• Stronger AI ethics frameworks, ensuring AI-driven software development is secure, compliant, and fair. The regulatory landscape around AI-generated code is evolving, and businesses will need to navigate stricter guidelines on transparency, accountability, and cybersecurity.
• AI’s influence on software engineering education and training, as universities and enterprises will need to reimagine developer skills for an AI-first era. Future engineers will need to collaborate with AI tools effectively, shifting their focus from repetitive coding tasks to higher-level problem-solving and strategic thinking. Platforms like DataCamp , Coursera , and GitHub Learning Lab are already adapting to this trend.

Gartner 's Hype Cycle gives us a fascinating look at the state of AI in software engineering, but the real impact happens in practice. Which trends do you think will dominate in 2025? Have you seen these shifts reflected in your own work?

Share your thoughts.