Client: "AI? That sounds risky for our project..."
Me: "you're already engaging with AI: the antispam filter.
Spam filters are AI at work, learning and protecting your inbox using your real data. You likely have given your consent at some points to have your emails processed by a third party, looking at the content to make a judgment on whether it is spam or not.
The core question is about trust: Who are you willing to trust to handle your data?
When you are using services that claim that all data can be read by a third party for 'research and improvement' purposes, then you should refrain from sharing secrets with this service. This is why you have to do your due diligence to crosscheck their claims, such as who is behind the service you intend to use.
There are two fundamentals if you want to start a project with AI:
1. Where do you send and store your data?
2. What is the company processing your data?
It is unlikely you will stop using emails in the near future. More sophisticated AI will be part of our daily life (Apple's announcements on automatic sorting is a good example); the question is, will you be ahead or just follow the flow.