To be fair agentic platforms are a shitty idea from a security perspective anyway. Everyone just threw that out of the window because shiny. Their funeral.
what's the worst that can happen?
Hackers trolling you? Stealing and deforming your projects? Hackers sharing personal info about you and your family and/or kids with the trashiest of people from your past or with your local authorities and wannabe matadors? All without your knowledge but you notice and get pseudo-paranoid and raise your security levels although it is technically and practically too fucking late?
Damn, how much these companies Anthropic, OpenAI, Microsoft, Meta, and ISPs and their hardware suppliers and all those parties and mayors and cities care about their paying customers and users and the people, workers and their children ...
Let's build more data centers and make AI infer smarter & faster. Better margin, better life.
Addendum: We have entered the era where companies carry more responsibility--if not all of it in tandem--for their users' digital security and privacy than the users themselves, especially those for whom the RAM et al prices are a tad bit much but who still pump considerable amounts of time, data and money into various services.
Social media convinced people to publish on the web details of their personal life that they never would have voluntarily handed over to federal agents: political views, personal relationships, pictures of everywhere they've been, graphs of associates.
AI gets people to interactively divulge what they're working on and even to upload the files. It's the next step on a path from a past where intelligence agencies had to spy on people to a future where people pay for the privilege of volunteering everything to the intelligence agencies.
I wasn’t familiar with the details but that’s not a very convincing connection to derive anything from. Is there more evidence to the story? Seems like a small amount of money and they briefly crossed organizational paths for about a year, over 20 years ago.
Amazon have very expensive lawyers, I'm sure that their T&Cs are flexible and woolly enough to include some reasonable justification for doing whatever they please with your VM.
"in order to secure our service against abuse..."
"in response to a court order..."
"...in order to provide better service, we reserve the right to."
https://aws.amazon.com/service-terms/
96. AWS Nitro System
AWS personnel do not have access to Your Content on AWS Nitro System EC2 instances. There are no technical means or APIs available to AWS personnel to read, copy, extract, modify, or otherwise access Your Content on an AWS Nitro System EC2 instance or encrypted-EBS volume attached to an AWS Nitro System EC2 instance. Access to AWS Nitro System EC2 instance APIs – which enable AWS personnel to operate the system without access to Your Content - is always logged, and always requires authentication and authorization. /quote
This has been third party validated (there's a blog) and is part of the contract terms. Which you would find if you did any research before posting. I'll let you figure out how you don't understand KMS either.
Good for them, if it's really the case that they can't read Nitro EC2 instances. Of course, that still leaves S3, RDS, DynamoDB, Elasticache...