Наталя ХандусенкоAI Eng
7 October 2026, 16:45
2026-10-07
"They try to learn everything about you, like a loved one": Meta's Muse AI agent accused of spying on users' friends and relatives
Meta's new AI agent Muse may have a fluffy and cute mascot, but behind that cute exterior lies a massive surveillance network controlled by one of the world's most powerful tech companies.
According to cybersecurity researchers, behind the attractive interface lies a large-scale data collection system that violates the privacy of not only the users themselves, but also those around them.
Meta's new AI agent Muse may have a fluffy and cute mascot, but behind that cute exterior lies a massive surveillance network controlled by one of the world's most powerful tech companies.
According to cybersecurity researchers, behind the attractive interface lies a large-scale data collection system that violates the privacy of not only the users themselves, but also those around them.
Despite rave reviews from reviewers, technical analysis of the tool revealed serious threats.
As Wired reports, Muse's AI agent keeps detailed profiles of all the people mentioned in a user's conversations and updates them regularly. According to the publication, the database is updated hourly, and internal instructions explicitly instruct the system to maintain "a separate page for each person in the user's life."
AI security researcher Karan Joshi, who extracted Muse’s internal prompts and shared them with Wired, highlights the deep level of data collection: “Judging from all these system prompts and the data that they’re feeding Muse, they’re trying to understand your relationships with real people,” Joshi said. “They’re basically trying to learn everything about you, like a close person, which is, frankly, scary.”
To make matters worse, just before launch, Meta's engineering team discovered a critical vulnerability that could have allowed users to gain direct access to the internal database that powers Muse.
According to 404 Media , the fixes were developed in a hurry, literally at the last moment before the release.
“Many senior engineers believe it is inevitable that we will have a large-scale data leak as a result of Muse,” a Meta insider noted in a comment to 404 Media.
It remains unknown at this time whether the formed dossiers on "friends and relatives" will be included in this database, but experts are unanimous: the launch of Muse creates unprecedented risks to the privacy of people who did not even consent to the processing of their data.
UPD Instead of an answer — a ban: Meta blocked the account of a Ukrainian IT entrepreneur after his request about restrictions on Muse models in Ukraine
«В жовтні випускаємо VR-шолом для аватарів, в «чіпування» Neuralink Маска вірю мало». Про що глава Meta Цукерберг 3 години говорив в подкасті Джо Рогана
25 серпня вийшла чергова серія популярного подкасту The Joe Rogan Experience, гостем якого став глава компанії Meta Марк Цукерберг. Розповідаємо про головне з майже 3-годинного інтерв’ю.
Як нейромережі бачать вільну та незалежну Україну? Тест dev.ua
Нейронні мережі для генерації зображень бачать світ по-своєму, їхню логіку зрозуміти часом зовсім неможливо. Але таки хочеться. На честь Дня Незалежності України редакція dev.ua вирішила провести невеликий експеримент.
Ми задали чотирьом різним нейронним мережам п’ять однакових запитів: «прапор України», «День Незалежності України», «український Крим», «перемога України» та «українці». Отриманими результатами ми ділимося з вами нижче.
Have important news to share? Message our Telegram bot
Key events and useful links in our Telegram channel
