NYT Opinion Says AI Won’t Take All Jobs After 20,000 Instagram Accounts Were Hijacked
Updated
Updated · The New York Times · Jun 30
NYT Opinion Says AI Won’t Take All Jobs After 20,000 Instagram Accounts Were Hijacked
1 articles · Updated · The New York Times · Jun 30
Summary
A New York Times opinion piece argues fears of an AI-driven job wipeout are overstated because large language models cannot reliably reason through many real-world tasks.
More than 20,000 Instagram accounts were reportedly taken over after scammers manipulated Meta’s AI-based support system, which the article cites as a glimpse of what fuller automation can look like.
The piece says similar failures hit other companies: Air Canada shut a chatbot after it wrongly promised a refund and lost in court, while McDonald’s dropped a drive-through bot after repeated breakdowns.
Its core claim is that LLMs are “plausibility engines,” not reasoning machines, so hallucinations and exploitable mistakes are built into how they work rather than temporary bugs.
That limitation, the article concludes, means AI may automate some functions but is unlikely to replace the vast majority of human jobs without costly failures.
Is the rise of AI creating a hidden workforce to manage its constant, embarrassing failures?
If scammers can easily fool corporate AI, what happens when it runs our critical infrastructure?
Instagram’s 2026 AI Support Tool Breach: 20,225 Accounts Hacked, Security Lessons, and the Coming AI Workforce Shift
Overview
In late May 2026, Instagram suffered a major security breach when attackers exploited a vulnerability in an AI-assisted support tool. By taking advantage of a flawed password reset workflow, hackers generated unauthorized reset links and hijacked user accounts. However, accounts with Multi-Factor Authentication (MFA) enabled were protected from this attack. Once Meta discovered the flaw, they quickly disabled the compromised tool and took steps to secure affected accounts. This incident highlights the risks of integrating AI into sensitive systems and the importance of strong security measures like MFA to protect users from emerging threats.