Ethical vs. Unethical, Pro-Privacy vs. Against Privacy are the two common discussion points. I, however, think the bigger problem here is that there's a very non-zero probability that this technology may cause unintended consequences simply by relying on false/inaccurate data.
For one, I work in analytics (loaded catch-all occupation) and I work with people who would marry their "data skills" if they could. In my industry, false positives of 80% is acceptable, and openly admitted errors in "machine-learning" logic (quoted to highlight my company's buzz-word usage, but practically non-existent) are made daily. People create algorithms, and people make errors.
Let's let our imagination run wild here for a second: It's 2030, and this technology becomes ubiquitous to the point where no one objects. Businesses take all the data from sentiments, gender, age...etc. to optimize for their target demographic, and price accordingly. In other words, let's assume this tech is used for perfect price discrimination. Economic theory dictates this is a win-win for everyone since everyone starts paying their willingness to pay. But, let's assume there's a catastrophe and medicine is in dire need. Price discrimination works fine assuming perfect competition, and is a useful framework, but it breaks down empirically where we live in a society that doesn't behave so rationally. Who survives? Those willing to pay the most, and the algorithm worked flawlessly here. But it was not intended to dictate who survives.
What I'm trying to say is that we should be cognizant of the fact that we don't live in a perfect bubble, and technology like this should be scrutinized for it's effects exhaustively- including any unintended consequences. We live in a society (duh), and as a society, it is up to us, with the help of policy makers, to determine the fate of this technology.