| So (getting my head round this) ... mobile app creators added an App Annie agent to their apps, which downloaded performance / usage data on their app to central servers. And App Annie promised not to sell that data to third parties unless it was "aggregated and anonymised". (Does App Annie pay the app creators for this?) App Annie sells this so people can work out which app to advertise on (the one with all the usage!) and which all to invest in (the one with all the usage). But they found that no-one wanted the aggregated data, or it was nit accurate enough, so they stopped using anonymised data and used the raw data. but said "it's ok we have permission" Ok. So they signed contracts on both sides, which contracts directly contradicted each other - they were always going to get caught. But this is aggregated app usage data - Apple could publish this in a heart beat and (presumably) it would be legal and App Annie would have no market. I know that "everything is securities fraud" and if you lie to both sides you will get caught. But this feels like a non-prosecution. The SEC had to - it was so blatent when it gets laid out, but really I doubt anyone thinks this will drain the swamp. It's all usage data. Edit: less ! marks |
> [App Annie] went to great lengths to assure [Trading Firms] that the financial and app-related data [App Annie] sold was the product of a sophisticated statistical model and that [App Annie] had controls to ensure compliance with the federal securities laws. These representations were materially false and misleading
It's illegal because App Annie was feeding trading firms insider information while swearing it was actually just a really good statistical model. Presumably the actual model wasn't good enough, so they started using unaggregated data to eke out perf.
What they did is not just breach of contract, but insider trading as they fed private information to trading firms.
> the order finds that from late 2014 through mid-2018, App Annie used non-aggregated and non-anonymized data to alter its model-generated estimates to make them more valuable to sell to trading firms.