Data Rights Are Not Bargaining Power
The right to refuse a use of your data matters. Whether that refusal changes the deal depends on what you can do next.
Imagine a service you use every week adds a proper data-sharing control. You can finally refuse a use of your information without abandoning the product. A month later, the subscription price goes up for everyone.
You have gained something real. You may also have less money left at the end of the month.
That is a hypothetical, not a report about a particular company. But it exposes a gap in a story I find appealing: give people control of their data and they will capture more of its value. There is a missing step between the control and the better deal.
A right can change the price without changing who sets it
A 2026 paper by Yashuang Wei, Guofang Nan and Hubert Pun offers a reason to be careful. Their analytical model compares digital markets with and without consumer control over data sharing, under monopoly and duopoly conditions.
Within the model, control rights lead firms to raise product prices. Consumer surplus can fall; consumers benefit when personalisation is highly valuable and privacy concerns are strong. The competitive effect is also conditional: rights can shift advantage towards the weaker firm in a duopoly, potentially reducing monopoly risk.
These are model results, not measured price changes following GDPR. They do not establish that privacy regulation generally makes people worse off. The publisher’s introduction makes the market conditions central to the findings.
What I take from this is a narrower warning: changing a person’s data choices need not improve their economic position. The surrounding market still matters.
The export button has to lead somewhere
Suppose you can download every record held by a service. The file arrives promptly, contains the right information and opens without specialist help. That would already be a decent product outcome.
Now try to use it somewhere else.
Perhaps the competing service cannot import it. Perhaps it accepts the records but loses the categories, corrections and years of context that made them useful. Perhaps there is no credible competing service at all. You keep the file and renew the subscription.
The export worked. You were still dependent on the service.
Bargaining power comes from having an alternative the other party has reason to take seriously. A person who can compare offers, carry useful history across and leave at a manageable cost is in a different position from someone who can only select a privacy preference inside the same account.
For a builder, I think the practical questions are quite ordinary:
- Can another product actually use the exported information?
- How much history or functionality does someone lose by moving?
- Can they compare the full price and service before committing?
- Can they get help exercising a choice without giving a new intermediary unrestricted access?
These are my product questions, not interventions tested by the paper. They concern the steps between having a right and being able to use it to change your circumstances.
Personal AI could help with some of that work. It could make a difficult export legible or prepare a comparison. It could also charge for access to the alternatives, favour its commercial partners or become another place where leaving means starting again. An assistant does not become an independent representative merely because it speaks in the first person.
Privacy does not have to pay for itself
There is a serious objection to treating any of this as a test of whether data rights are worthwhile.
People have reasons to protect their lives that do not fit comfortably into a calculation of price and service quality. A person may want to explore a health concern, change their beliefs or keep an intimate relationship outside a commercial profile. They should not have to demonstrate that doing so produces an economic gain.
The European Data Protection Supervisor describes privacy and data protection as supporting fundamental values and other freedoms. That is a broader purpose than getting consumers a cheaper subscription.
Dignity and autonomy are not exhausted by the preferences represented in a market model. Nor should willingness to pay decide who deserves a protected private life. Someone with very little spare money can care deeply about privacy and still be unable to afford a premium tier.
Higher prices also need interpretation. Real security, compliance and protection work can cost money. A price increase alone does not tell us whether a firm is recovering those costs, exploiting market power or doing some combination of the two. We would need evidence about the particular market and business.
The useful response is to hold both questions open: are people properly protected, and can they obtain a better deal? Success on one does not settle the other.
What happens after someone says no?
I still want people to have control over the information their lives produce. I also want the product conversation to continue after the settings screen is finished.
If someone refuses a use of their data, can they keep using the service on fair terms? If they leave, can the next provider use the history they bring? If neither route is practical, who can help them challenge the situation?
A working control is worth building. So is the path out of the room.