Hackers tricked Instagram AI into letting them take over 20,000 accounts [U]
Hackers managed to trick Meta’s AI-powered support bot into allowing them to take over a number of Instagram accounts, including some high-profile ones. This included accounts belonging to the White …
![Hackers tricked Instagram AI into letting them take over 20,000 accounts [U]](https://9to5mac.com/wp-content/uploads/sites/6/2026/06/Hackers-tricked-Meta-AI-into-letting-them-take-over-high-profile-accounts.jpg?quality=82&strip=all&w=1600)
Hackers managed to trick Meta’s AI-powered support bot into allowing them to take over a number of Instagram accounts, including some high-profile one
Read Full Story at 9to5Mac →Why This Matters
The breach exposes a critical vulnerability in Meta’s AI-driven security infrastructure, demonstrating how automated systems can be manipulated at scale. Beyond the immediate risk to user data, this incident underscores the growing sophistication of cybercriminals in exploiting loopholes in corporate AI defenses—raising urgent questions about the reliability of AI as a standalone safeguard in high-stakes digital environments.
Background Context
Meta’s AI support bot, designed to streamline user authentication and recovery processes, has increasingly relied on machine learning to detect fraudulent activity. However, social media platforms have long been prime targets for credential stuffing and social engineering attacks, with Instagram—a hub for influencers, businesses, and public figures—often serving as a high-value prize. The platform’s rapid user growth has outpaced the scalability of its security protocols, leaving gaps that adversaries can exploit.
What Happens Next
Meta will likely patch the AI’s decision-making logic and introduce stricter human oversight for account recovery requests, but the damage to trust may linger. Regulators may also scrutinize the incident as part of broader discussions on AI accountability in platform governance. For users, the takeaway is clear: relying solely on automated systems for security is no longer sufficient—multi-factor authentication and vigilance against phishing remain essential.
Bigger Picture
This breach reflects a troubling trend where AI systems, deployed to reduce operational friction, inadvertently create new attack surfaces. As platforms prioritize automation to manage vast user bases, adversaries are shifting their focus to exploiting these very systems. The incident also highlights the arms race between cybersecurity defenses and offensive innovation—a dynamic that will only intensify as AI becomes more deeply embedded in digital infrastructure.
