Conclusion

The primary objective of this project was to identify the micro alignment gap in AI safety. Specifically, how emerging technologies impact employee well-being, security, and privacy.

To meet this objective, an ethical governance framework and project management plug-in were created.

The solution was shown to be successful as it met set goals by scanning tickets in Jira for any incompliances, ensuring that AI development does not veer from human-centric design.

The project provides significant value to TD by offering a structured way to implement AI which, from our initial research, TD currently lacks. For the intended user, our solution reduces the cognitive load required to read and interpret a complex framework by having software do it. From this, the system empowers employees to make informed, ethical decisions during the design process.

On a broader scale, this solution proves that transparency and guidance are more effective than surveillance for maintainability. This demonstrates that AI can be a main driver of SDG 8 and SDG 3 by enabling a workplace where human creativity is backed by powerful technology. Sustainable development was kept in mind as the solution was built upon a modular architecture.

To maintain its effectiveness, the framework requires regular backups and maintenance. Its integration into standard tools like Jira ensures it can be scaled across various departments within TD Bank beyond the scope of this course.

While the MVP is functional, future iterations should focus on enhancing the suggestion engine to better capture the subtle contextual nuances of specific user requests. Based on our iterative refinement process, there is also an opportunity to further redesign the interface to reduce complexity even further and expand the violation detection modules to cover a wider array of emerging global AI regulations.

The underlying approach of this Micro-Alignment framework is highly adaptable to any industry undergoing rapid digital transformation, such as healthcare or education. Because the system prioritizes human-centered design and transparency, it could be extended to serve as a universal “ethical auditor” for any project management environment where human autonomy must be protected from algorithmic overreach.