Hacker News new | ask | show | jobs
by rokizero 745 days ago
It would have been fair if the author would have mentioned that Microsoft is very open about this. I went to a Microsoft training, where the instructor also made clear that the service is being monitored. Suspicious messages get flagged and are reviewed under a 4-eyes principle. [1]

At least in the EU (I'm told) they are required by law keep logs to ensure that their AI services are not being used to do bad.

I'm glad that their system worked.

[1] https://learn.microsoft.com/en-us/legal/cognitive-services/o...
2 comments
dkobia 745 days ago
They are indeed very open about this. More here [1]. You can also put in a request for limited monitoring [2].

[1] https://learn.microsoft.com/en-us/azure/ai-services/openai/c...

[2] https://customervoice.microsoft.com/Pages/ResponsePage.aspx?...

link
Hizonner 745 days ago
Very impressive. How many actual users waste days of their lives in Microsoft training to get basic information?

And, no, burying it in the ToS isn't being "very open" either. "Very open" would be putting a big visible banner on every page of the UI.

... and the abuse still wouldn't be acceptable even if Microsoft actually were being forthcoming about it.

link