The fairness edge of algorithmic pricing: personalization isn't a licence to discriminate

Algorithms can price dynamically by user profile, but when they systematically show different prices to different groups, they may cross price-discrimination, consumer-protection, and even antitrust lines.

Risk rises sharply when pricing differences correlate with protected characteristics (such as region or group), and 'personalization' rarely justifies it.

The pragmatic approach is to examine whether pricing logic is explainable, whether it systematically disadvantages specific groups, and to keep the pricing basis for challenges.