IT security help with AI
This is a future plan. IT security help with AI is not available in the mailbox today. The assistant is not switched on. If it existed, the person would still be the one who sends.
What it is
The plan is an assistant inside the mailbox that would do two narrow things. It would point at a sender that looks suspicious, so a person can look twice before they reply or open a file. It would help draft a reply, as text the person can read and change. It would not decide for them, and it would not act in their place.
It is for a person reading their own mail, and for someone at a company doing the same with a mailbox they are allowed to open. It is a second look and a draft, for people who want that help and still intend to judge the letter themselves. It is not a security team for hire, and it is not a product that answers the mail while the person is away.
Nothing in the current mailbox is that assistant. Mail is not being sorted by an AI feature. No model is reading the inbox in order to arrange it, warn on it, or write back. The plan names a future assistant. The name is not a switch.
How the logic would work
On a suspicious sender, the logic would be a mark, not an action. The assistant would look at the sender and the shape of the message and point when something looks off: a name close to one the person trusts, a request that hurries them, a file they were not expecting. The message would stay where it is. The person would open it or leave it. The assistant would not delete it, would not move it in secret, and would not reply.
On a draft, the logic would be a suggestion in the reply box. The person would ask for a draft, read the words, change them, and only then send. If they never press send, nothing leaves. If they close the draft, the suggestion is gone. The assistant would not keep sending attempts of its own.
The person would still be responsible for the address they send to and for the secrets they type. The assistant would not be a reason to paste a password, a Turkeuro account code, or a verification code into a message. A draft that asked for that would be a draft to delete, not a draft to trust. The plan would not train the person to hand secrets to a box.
Because the assistant is not switched on, none of this logic is running. A message that looks urgent is still a message the person reads with their own eyes. The absence of a mark is not an assurance. There is no mark.
What you would see
If the plan were on, a person would see a mark next to a sender the assistant found suspicious, with a short reason in ordinary words. They would see the message still sitting in the mailbox, unsent and undeleted. They would see a draft only after they asked for one, and they would see the send control still waiting for them.
They would see their own edits in the draft. They would not see the message leave while they were still reading. They would not see a claim that the assistant had already replied, because it would not reply.
Today they see the sender's address and the letter, with no assistant beside them. Login settings do not contain an AI control. There is no model to turn off, because none is on.
What it would not do
It would not delete mail. It would not file mail on its own. It would not send, forward, or reply. The person still sends. That sentence is the limit of the draft, and it is the limit of the warning.
It would not be a reason to trust a page that only looks like Turkeuro. The account door is accounts.turkeuro.com. The mailbox is mail.turkeuro.com. An assistant in a future mailbox would not change those hosts, and a mark inside a letter is not a substitute for reading the address bar of a site.
It would not ask the person to turn off their password or their saved sign-in choices. It is not switched on. A message that claims an assistant has already secured the account is not this plan, and it is not Turkeuro.
What is available today instead
You read the sender yourself. You decide whether to open the file and whether to reply. The mailbox gives you the letter. It does not give you an assistant.
The sign-in choices you can save are the password and, if you added them, the account code, a two-factor choice, an authenticator, and email as a verification choice. None of those is an AI feature. They are also not asked again at sign-in.
What you can use today is a mailbox on Postfix, Dovecot, and SnappyMail. You sign in with a password. The account can hold a 41-character Turkeuro account code that starts with turkeuro, a two-factor choice, an authenticator, and email as a verification choice. Those second-factor choices are saved with the account. They are not asked again at sign-in. The connection uses TLS. Login settings hold the sign-in controls for the mailbox you have open. There is a Windows desktop app. Business plans are listed at https://turkeuro.com/business.
SMS sends nothing. A mobile number saved on the account does not produce a text message from Turkeuro. The unlock choice in Login settings does not lock the screen. Those are limits of the mailbox as it is. They are not a quiet version of the plan on this page.
