Let ’em cook: Gen Alpha slang stumps AI moderation systems 92% of the time

A new study reveals that Generation Alpha’s rapidly evolving internet slang is creating blind spots for AI content moderation systems and adults trying to protect young people online. Research conducted by Manisha Mehta, a 14-year-old student, and Fausto Giunchiglia at the University of Trento, Italy, found that while 92% of Gen Alpha users can detect harmful intent in coded messages, AI models only catch about 40% of cases—and parents perform even worse.

What you should know: The research analyzed 100 popular Generation Alpha expressions from gaming and social media platforms, testing comprehension across different groups.

• Among 24 volunteers aged 11-14, 98% understood basic meanings, 96% grasped contextual usage, and 92% could identify harmful applications.
• Four major AI models (GPT-4, Claude, Gemini, and Llama 3) detected harmful use in only 32.5% to 42.3% of cases.
• Parents and professional content moderators performed similarly poorly, spotting only around one-third of harmful instances.

The coded language problem: Many Gen Alpha phrases carry double meanings that can mask bullying or abusive intent depending on context.

• “Let him cook” can express genuine praise in gaming streams or serve as mocking dismissal implying someone is talking nonsense.
• “Kys” has evolved from meaning “know yourself” to often signaling “kill yourself” among some users.
• “Is it acoustic” functions as a veiled way to mockingly ask if someone is autistic.

Why this matters: The research exposes critical vulnerabilities in online safety systems as AI-powered content moderation becomes more prevalent.

• “Gen Alpha is very vulnerable online,” says lead researcher Mehta, who attends Warren E Hyde Middle School in Cupertino, California.
• “I think it’s really critical that LLMs can at least understand what’s being said, because AI is going to be more prevalent in the field of content moderation, more and more so in the future.”

What experts are saying: The findings highlight fundamental gaps in current protection systems for younger users.

• “Empirically, this work indicates what are likely to be big deficiencies in content moderation systems for analysing and protecting younger people in particular,” says Michael Veale at University College London.
• “Companies and regulators will likely need to pay close attention and react to this to remain above the law in the growing number of jurisdictions with platform laws aimed at protecting younger people.”

The research approach: The study combined teenage insights with academic rigor to understand this communication gap.

• Mehta collaborated with Giunchiglia, collecting expressions from popular platforms used by those born between 2010 and 2025.
• “I’ve always been kind of fascinated by Gen Alpha language, because it’s just so unique, the way things become relevant and lose relevancy so fast, and it’s so rapid,” Mehta explains.
• The findings were presented at the Association for Computing Machinery Conference on Fairness, Accountability and Transparency in Athens, Greece.