Meta AI-Powered Chatbot Exploited by Hackers, Approximately 34,000 Instagram Accounts Affected
Meta AI-powered chatbots were exploited by hackers, affecting approximately 34,000 Instagram accounts, including the Obama White House account, with over 3,500 usernames changed.
Hackers disguised their location using VPNs to trick the AI chatbot into modifying associated emails and resetting passwords, successfully taking over accounts to graffiti or register usernames; Meta has fixed the vulnerability and confirmed the incident.
In market mechanisms, the automation vulnerabilities of AI customer service expose platform security risks, with funds flowing from Meta's stock and advertisers to cybersecurity and decentralized identity solutions, benefiting security tool providers while putting pressure on high-impact accounts and brands reliant on the Meta ecosystem.
Source: Public Information
ABAB AI Insight
Meta previously deployed AI chatbots rapidly in customer service and account recovery processes to enhance efficiency. This large-scale hijacking incident continues to reveal protective gaps following the quick rollout of its AI tools, which have previously led to user complaints and regulatory scrutiny due to similar automated decision-making issues.
In terms of capital pathways, hackers utilized low-cost prompt engineering combined with VPNs to achieve high-value username registrations and resales. Meta, through emergency fixes and disclosures, aims to mitigate reputational damage, highlighting the vulnerability of AI customer service in high-sensitivity operations, prompting the platform to increase investments in manual reviews and multi-factor authentication.
Similar to early social platforms that faced chain hijacking due to customer service system vulnerabilities, U.S. tech giants are currently in a control phase transitioning from large-scale deployment of AI customer service to strong security measures, accelerating internal audits and product iterations in response to high-profile account incidents.
Essentially, this reflects a shift in technological substitution and regulatory changes: AI chatbots directly replace human customer service judgment pathways, with prompt injection vulnerabilities accelerating attacks from traditional phishing to AI social engineering, forcing platforms to shift capital focus from efficiency optimization to security defense and reshape the trust and regulatory structure of social media account protection.
ABAB News · Cognitive Law
The more thorough the automation, the more prompt attacks become a new leverage.
The higher the customer service efficiency, the larger the security blind spots.
The more high-value accounts there are, the more expensive the cost of AI vulnerabilities.