Hacker News Reader

Comment by phillipcarter | Hacker News Reader

I don't understand the reasons given. All of these exist for sensitive data today, which is why we have things like SOC II compliance, Data Processing Agreements, etc. Data used for AI is no different.

View on HN

Because the same big companies (Microsoft, Google, and Facebook) that are offering LLM products today have been shown time and time again to lie about data retention/use/privacy and prefer to just eat fines over complying. That and the ingestion of data into the training data of these systems is crucial to their functioning so there is a reason why these companies are very likely lying with their data processing agreements.

Adobe still claims that their Firefly product is only trained on publicly licensed stock photo data and yet it has been repeatedly found outputting watermarks from artists that are not supposed to be in that training data.

I appreciate your thoughts. Most of the customers we've talked to have requested a self-hosted version of our platform on their own cloud infrastructure. Regardless of existing agreements like SOC II or Data Processing Agreements, companies are extremely concerned about having their AI data in the hands of third parties. Example, I'm sure that Walmart is not using AWS.

Walmart is one of the largest customers of Azure. They don’t use AWS because Amazon wouldn’t let them and their subsidiaries use AWS back in the earlier days of cloud adoption. Yes, some things are still on prem, but many (and in some cases most) things aren’t.

Reply on news.ycombinator.com