Summary video - AI Powered Threat Detection
Apologies in advance, I have the fashion/design sense of a rock. My boss is ok with that, and I hope you are too.
My socio-technical plan involves implementing AI-powered threat Detection at my current employer, Riverbed. The video I created, built on top of a PowerPoint presentation, goes over the objectives, how I envision implementation going, the challenges I can see arising, how to best execute stakeholder engagement, the ethical concerns involved, and how to best evaluate and monitor the system so that future improvements can be implemented. It was more complicated than I care to admit to getting my Individual Project boiled down to 8 slides, but I could do it. Something I wanted to find a way to insert into my plan but wasn’t able to was to find a way to integrate this plan into our products. While that would have been a lot of fun to explore, it would have been WAY more complex and challenging, so I focused on implementing this internally, facing our infrastructure. I envision that once executed internally, finding a way to integrate this into our products with the lessons learned would not be that difficult.
That said, there is significant value in rolling out this plan internally first. It allows us to refine our approach and iron out any potential issues and ensures that we have a robust and secure environment before considering product integration. This approach provides a solid foundation, enabling us to leverage our internal successes as a proof of concept for future product enhancements. Also, by focusing on our internal infrastructure, we can better tailor the threat detection system to meet our specific needs, setting the stage for future integration into product lines. Ultimately, this approach mitigates risk and maximizes the potential for successful implementation and long-term sustainability without burning out the small security team. Resolving the political problems that will invariably arise will undoubtedly be challenging, seeing as various teams will want to limit access to some data and control the outputs of other functions. Still, the reality is that for a productive output of the plan, as much as possible needs to be included and used by the AI system. Strong leadership will be required to ensure the program's success, guiding it past these potential issues.
Comments
Post a Comment