As artificial intelligence becomes an everyday tool for companies and consumers alike, one thing is clear: making ethical choices isn’t optional anymore. Whether you’re a developer, product manager, or business leader, treating ethics as an engineering challenge will help you build solutions that benefit people—and your company—in the long run.

In this blog post, we’ll explore four key ethical choices to make across three critical areas: Data Source, AI Lab/LLM, and Effects on End Users. Let’s look at these areas and practical considerations to guide you.

The Challenges

1) Data Source

The data you use can hold stolen content, biased datasets, or personal information gathered without proper consent. It’s vital to ensure all data is acquired fairly, with clear permissions, and that you understand the inherent biases present. Consider the human labor that went into labeling and normalizing this data and whether workers were fairly compensated. Also reflect on the environmental impact of ingesting huge datasets into your system.

2) AI Lab/LLM

When sourcing a powerful model or partnering with an AI lab, make sure their training pipeline respects data rights. Assess their parameters, weights, and bias controls, and check whether their company culture reflects your personal ethics. Be comfortable with the full story behind the model, knowing customers may eventually want to understand how it was created.

3) Effects on End Users

Your AI-driven product can cause unintended harm if you haven’t thought it through. Consider harmful outputs like hallucinations that could cause real-world damage, and acknowledge any risks to jobs or lives. Have a plan for responding to customers who’ve been negatively affected and strive to ensure the impact aligns with both your personal and company values.

To navigate these challenges, engineering ethics must become part of the process—not an afterthought. Audit your data sources thoroughly, understanding where the data came from and acknowledging any existing biases. Verify the ethical guidelines of your AI labs and vendors to ensure you trust the people creating the models. Test for impact and bias continuously to catch unintended consequences and involve diverse stakeholders in testing. Finally, own your responsibility by planning, designing, testing, and iterating until your impact is one you can stand behind with pride.

Building ethical AI requires making thoughtful choices every step of the way—from your data to your model to the people who use your product. Treat ethics as an engineering challenge and you’ll reduce risk, build trust, and create tools that reflect your values. Beyond company goals, your decisions shape the future of technology and society. Take a moment to reflect on your role, advocate for transparency, and hold yourself and your colleagues accountable. By choosing responsibility and integrity every day, you help ensure that AI benefits everyone and earns the trust we all depend on.