Read parking data licences alongside privacy promises to understand model training, analytics, de-identification and uses after exit.
The agreement should describe permitted analytics, model development, disclosure and commercial uses, including what continues after exit. Review those terms alongside the privacy policy, with attention to the actual inputs and whether people remain identifiable.
A statement about not selling personal information does not answer every secondary-use question. Model training or product improvement can be another use of customer records even where no sale takes place.
An individual parking record serves a visit or review task. Combining records creates a dataset useful to analytics or development teams. That change introduces another purpose, rather than merely a new way of displaying the original operational record.
Replacing a plate with a code can still leave a detailed movement pattern. De-identification needs assessment against the detail retained, the recipients and other information available to link a trace back to someone.
A planning question may need counts by broad time period rather than individual histories. Compare that narrower report with the proposed dataset before agreeing to secondary processing, while still considering whether the aggregate output could identify someone in context.
A daily total for a busy zone differs from one coded vehicle appearing at a small site every morning. Removing the plate from the second example does not necessarily make the remaining record anonymous.
Establish who receives the inputs, whether retained customer records are included and what restrictions apply after termination. New purposes and changes to subcontractors need a defined assessment process, rather than an assumption that an earlier operational approval covers them.
Commercial terms should distinguish personal information from non-identifying outputs. A no-sale statement about personal information should not be presented as a perpetual prohibition on all commercial use of aggregate statistics.
Aero Ranger commits not to sell personal information while allowing specified service and legally authorised disclosures. Its agreement licenses customer data for model training, evaluation, improvement, analytics and research, using de-identified or aggregated inputs where practicable.
The licence can continue while data is held after exit. Deletion does not undo models or derived data already created. Procurement therefore needs to examine the intended purposes and inputs, without assuming anonymous-only training or a ban on all commercial uses.
Aero Ranger: Privacy policy; Aero Ranger: master services agreement
Sources were reviewed on the publication date. Check the linked material for current details.
