Artificial Intelligence (AI) is transforming the landscape of software engineering.

At PyCon APAC 2025, industry leaders discussed how AI is reshaping the field, the challenges it presents, and how executives and engineers can prepare for the future. Here’s what we learned from the panel featuring  Dominic Ligot, Jeremi Joslin, Iqbal Abdullah, and Younggun Kim.

AI isn’t replacing developers—it’s making them more efficient. Tools like GitHub Copilot and AI-assisted development frameworks are automating repetitive tasks, accelerating software delivery, and even enabling non-engineers to contribute meaningfully to development.

1. AI as a Junior Developer

• AI functions as an automated coding assistant, generating code, suggesting improvements, and running tests, but human oversight remains essential.
• AI-assisted pair programming can help refine code quality, but it doesn’t replace critical thinking and structured problem-solving.
• Companies leveraging open-source models are accelerating development cycles and making software more accessible to broader teams.

Case Study: At Shopify, AI-driven code completion has cut development time significantly, allowing engineers to focus on complex business logic instead of boilerplate code.

2. The Need for Thoughtful Engineering

Despite AI’s capabilities, strong software fundamentals are still indispensable. AI can generate solutions, but it lacks reasoning. Engineers must:

• Engage in code reviews to ensure AI-generated code aligns with business goals and best practices.
• Understand software architecture, as the emphasis shifts from writing code to designing scalable and maintainable systems.
• Ask 'why' at every step, ensuring that AI-driven automation aligns with business objectives rather than blindly trusting generated outputs.

The Risks of AI in Software Engineering

AI adoption comes with critical risks—ranging from security vulnerabilities to biased decision-making. The panelists emphasized the need for accountability and governance in AI-driven software development.

1. AI Bias and Security Risks

• AI models are trained on publicly available data, making them susceptible to bias and ethical concerns.
• AI-generated code has introduced security vulnerabilities, as demonstrated in research where fine-tuned models injected exploitable flaws into software.

Three major AI risks:

1. Misuse – AI tools can be leveraged for harmful applications, from misinformation to cyber threats.
2. Functionality Risks – AI models can generate hallucinated or incorrect outputs without clear accountability.
3. Systemic Risks – AI-driven decision-making can reinforce existing biases in hiring, lending, and other critical domains.

Case Study: Amazon discontinued an AI hiring tool after discovering it disproportionately favored male candidates, illustrating the hidden dangers of unchecked AI automation.

2. AI Requires Human Oversight

AI is only as good as its governance. Developers and executives must prioritize transparency, validation, and risk mitigation in AI adoption:

• Set clear AI guardrails to define acceptable AI-driven decision-making.
• Use AI for automation, but ensure that human validation is part of the process.
• Prioritize AI observability, tracking how models evolve and ensuring they remain aligned with company goals and ethical considerations.

3. AI is an Iterative Process

AI systems do not behave in a deterministic way. They require ongoing monitoring, refinement, and strategic integration:

• AI models change over time, meaning performance and accuracy need continuous evaluation.
• Fine-tuning AI systems to align with business needs is an ongoing process.
• The role of engineers is shifting from coding to orchestrating AI workflows, managing AI-driven automation in a structured, business-aligned manner.

Preparing for the Future of AI in Engineering

AI has seen over $1 trillion in global investment, making it a central focus for businesses looking to innovate. Businesses that successfully adopt AI will see accelerated innovation and efficiency. The panelists emphasized:

1. Experimenting with AI – Companies should actively test AI’s potential in workflows, automation, and software delivery.
2. Building AI literacy within teams – Understanding AI’s capabilities, limitations, and risks will be a competitive advantage.
3. Aligning AI adoption with strategic goals – AI should be a business enabler, not just a technical experiment.
4. Fostering AI education – Companies must invest in AI training programs, workshops, and upskilling initiatives to prepare teams for the next phase of AI-driven software development.

Key Takeaways for Executives

For executives evaluating AI integration, the key lessons from PyCon APAC 2025 include:

1. AI is a force multiplier, not a replacement. AI-enhanced development is about acceleration, not automation without oversight.
2. AI trust must be built through transparency. Observability, explainability, and security must be built into AI-driven software.
3. AI investment should drive long-term competitive advantages. Companies that integrate AI effectively will find new opportunities for growth and efficiency.
4. AI adoption requires adaptability. Businesses must continuously refine their AI strategies as technology evolves.

How Swarm Can Help

At Swarm, we specialize in helping businesses navigate AI adoption with a focus on real-world impact. Our expertise spans across:

1. AI Strategy & Implementation – We help organizations integrate AI solutions that align with business goals, ensuring efficiency without compromising security or ethical considerations.
2. Custom AI Solutions – Whether it's AI-assisted development, workflow automation, or AI-powered decision-making, we tailor solutions to your unique needs.
3. AI Education & Upskilling – We provide training, workshops, and consulting to help teams build AI literacy and develop best practices for AI-driven software engineering.
4. AI Governance & Risk Management – Our approach includes transparency, observability, and security frameworks to ensure AI is implemented responsibly.

Partnering with Swarm means equipping your teams with the right AI tools, strategies, and frameworks to stay ahead in an evolving digital landscape. Let’s build AI solutions that drive business impact together.

Conclusion

AI is becoming an essential tool in software engineering, helping teams build better software at a faster pace. Executives and engineers alike must focus on understanding AI’s capabilities, ensuring ethical adoption, and maintaining strong software engineering principles.

At Swarm, we’re committed to building AI-powered solutions responsibly, focusing on ethical adoption, thoughtful engineering, and sustainable business impact.

How is AI shaping your software development process? Let’s talk.

‍