The Power of AI, but a Word on Privacy

AI is incredibly powerful and indeed with a client we recently used DeepAgent and other platforms to help us rank two preferred candidates for a senior executive position. AI, instantaneously, produced both an informative graphical summary but, I must say, also an extremely good and succinct summary of the candidate’s relative strengths. We input our candidate reports and also interview notes. I couldn’t have summarised it better myself!

It was, as I say, a powerful example of the benefits of AI and something I would certainly use again.

However, it did occur to me that we must be ever vigilant with the data we put through AI. There are privacy concerns as AI relies on vast amounts of personal data. Indeed, its ability to collect, process and analyse vast amounts of sensitive information. We should, when using AI, protect the names of individuals as much as possible – give them made up names! After all, we obviously cannot hold AI accountable for any breaches!

There are concerns that stem from security, the potential mis-use of information and a lack of transparency. AI systems can be victims of cyberattacks, whilst AI models are trained on data and that data reflects societal biases and AI may perpetuate or amplify that bias. I also feel that the “black box” nature of some AI Systems can make it difficult to fully understand how it makes some decisions or rankings.

Regulations around AI and privacy obligations are still evolving and we need to be careful. There are some immediate measures we can make. In Chat GPT for instance, in Settings, under Data Controls, you can elect to ‘opt out’ of the default “make model for everyone”, this prevents Chat GPT from “training itself” with that data. There are also ‘Temporary’ chats that can be used for sensitive work.

However, without doubt, when treated “with care”, AI is an incredible agent for productivity but also undoubtedly also for value added information and advice.