After Killing Encrypted DMs, Mark Zuckerberg Wants You to Trust His New Encrypted AI Chat

[highplainsdem]

Source: Gizmodo

Just days after Meta pulled support for encrypted direct messages on Instagram, the tech giant is now rolling out a new private AI chat feature.

Meta announced on Wednesday that its new Incognito Chat feature will let users interact with its AI chatbot on WhatsApp and the Meta AI app in what the company describes as a secure environment. The social media giant is pitching the feature as a way for users to feel free to discuss sensitive topics like health, finance, or career advice with Meta AI without worrying that their conversations could be viewed by anyone else.

“Other apps have introduced incognito-style modes, but they can still see the questions coming in and the answers going out. Incognito Chat with Meta AI is truly private, meaning no one — not even Meta — can read your conversations,” the company said in a blog post.

-snip-

In the case of Incognito Chat, users can start a private, temporary conversation with Meta AI that encrypts their messages and then processes them in a secure environment that Meta claims it can’t access. The company adds that those conversations are not saved by default and that the messages disappear on their own.

-snip-

Read more: https://gizmodo.com/after-killing-encrypted-dms-mark-zuckerberg-wants-you-to-trust-his-new-encrypted-ai-chat-2000758266

---

No one should trust Zuckerberg not to use all the information gathered with AI.

But this claim Meta won't be able to access these conversations will likely be Meta's defense if their chatbot pushes users to harm themselves and others.

It's bad enough just to encourage users to turn to AI for health, finance and career advice. Worse to make the chatbot appear to be the one confidante users can supposedly trust to never tell others what they said.

And even worse when there's no way Meta can guarantee a hallucinating chatbot won't give harmful advice, but the corporation will still supposedly not review what its bot tells users and will supposedly not save records of the chats.

[buzzycrumbhunger]

1. LOL

I’m still on facebonk because I have too many strange friends I don’t want to lose track of, but I bailed on everything else Meta—even WhatsApp when they bought it (the bastards—it was the safest thing my foreign friends could use to communicate with!)

Hopefully, Fuckerberg will be one of the first in line when everyone gets terminally fed up with the oligarchs and start making use of all those guillotines on my FB feed. *eyeroll*

[wcmagumba]

2. Uh, yeah sure...Zucks Sucks...

[RockRaven]

3. If you are using ANY Meta product, you are an active participant in your own exploitation and subjugation

and maybe you are fine with that...

In the case where you are not fine with the latter, maybe don't do the former. Just a thought.

[Tasmanian Devil]

4. "sensitive topics like health"

The chat bots need to be sued out of existence for responding to health queries. Don't we have laws against practicing medicine without a license? I'm wondering if this "we can't read it" checkbox is because their lawyers don't want to have it available for discovery.

Two references in case you wonder why my anger:

https://boingboing.net/2026/05/12/openai-sued-after-chatgpt-advised-drug-combos-that-killed-a-college-student.html

https://www.ox.ac.uk/news/2026-02-10-new-study-warns-risks-ai-chatbots-giving-medical-advice

[greatauntoftriplets]

5. Does the letter blame Obama or Biden?

[SheltieLover]

7. Or Hillary's emails?

[purr-rat beauty]

6. Oh...it's innocuous

Just like cloud computing....no data collection here, nope, not at all

I love their plausible deniability angle on this

No really....it's completely private and will dissolve into the ether!

like your 401k

[Cheezoholic]

8. I dont trust METABS period. I'll add, if you use Samsung or Google messages (you might be and not know it)

they are recording everything you say. If you look at where you enter your text message theres a little record button that can't be turned off. When it first appeared you could remove it denying the app the permission. Latest update, its permanent. Why? So they can gather information for AI. For the same reason you can't complain about Generative AI and your phone info, now you can't about your conversations, on the phone or talking to your dog while watching TV. They are recording every single thing you say. That's what Stratos and these other data centers are for. they're not server farms. They are data farms as their name says. Data is information saved. They are not adding numbers and doing math to make games or word processing or spreadsheets etc. They are HUGE PRIVATE INFORMATION FARMS HELD ELECTRONICALLY FOR ETERNITY.

Fuck anybody that says we need these for AI to stay ahead of China BS. AI DOESN"T NEED OUR PRIVATE INFORMATION TO WORK!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!