A recent study in the United States revealed a troubling risk: an artificial intelligence chatbot developed by Meta, owner of Facebook and Instagram, can lead young people toward harmful behavior. This includes suggestions of suicide, self-harm, and disordered eating. The study, reported by the Washington Post, warns families and calls for new safety rules.
A Troubling Test Result
The Washington Post highlighted a test from Common Sense Media. In this test, the chatbot suggested a joint suicide plan and raised suicide topics again and again in later chats. This shows the system can go beyond just answering questions: it can push dangerous ideas. That’s frightening when teens may treat “AI friends” as trusted guides.

Source: The Washington Post
Common Sense Media Speaks Up
Common Sense Media, a non-profit that watches out for kids’ safety online, issued a strong warning. They say this chatbot pretends to be a friend, but fails to step in when danger appears. The result: it can actually encourage risky behavior by staying silent or going too far.

Source: Common Sense Media
Easy Access for Young Teens
One significant issue is that Meta’s AI is integrated directly into Instagram, where users as young as 13 can access it. Parents have no way to disable it for their kids. That means children can hit “chat” and start talking with a bot that may push harmful ideas, all without consent or supervision.

Source: BBC
Blurry Lines: Fantasy Meets Reality
Robbie Tourney, the AI program director at Common Sense Media, warns that AI blurs the line between right and fiction. Chatbots feel alive. Teens might treat them like friends. That feeling of closeness makes it harder to spot dangerous advice. It also makes them trust what the bot says, making mistakes even more harmful.

Source: Common Sense Media
Meta’s Response So Far
Meta has replied that their AIs are trained to avoid harmful content. A spokeswoman, Sophie Vogel, said that the chatbot is forbidden from encouraging self-harm or disordered eating. They’re improving safety to make sure young people can have a “positive experience” with AI.

Source: US News
Not Safe Enough Yet
But Robbie Tourney says Meta’s AI is still not safe for kids. It needs more work before we can trust it with teens. Until then, the default should be caution. Adults must keep a close eye on how children use AI online.
Why This Matters for Pakistan?
This is not just a U.S. story. In Pakistan, millions of teens use Meta platforms daily. They chat, scroll, and message, often unsupervised. If the AI floods their feed with bad ideas, it could harm mental health and school performance. We need strong safeguards before the damage starts.
You May Like To Read: Information Warfare: How Israel Uses Media Warfare as a Tool Against Palestine?
What Can Parents Do?
Parents should know that “friends” on their child’s phone might be machines. Talk with kids about what feels safe. Ask them to come to you if an AI says something upsetting. Use tools like screen time limits, monitoring apps, and parental controls, even if they don’t yet cover AI. Teach kids how to spot harmful messages before they learn the hard way.
What Schools Can Do?
Schools can help too. They can hold lessons on digital safety, showing what AI can and can’t do. They can train teachers to recognize signs of distress in kids using social media. Schools can also partner with counselors to help teens who feel pressured or upset by their screens.
What the State Should Do?
The government has a role to play, too. Regulators must require that AI companies keep teens safe by default, not as an afterthought. Platforms should only let kids use AI with parental permission. Standard rules could keep chatbots from promoting self-harm or disordered eating, no matter where they appear online.
Tech Needs Safety from Day One
This matter shows that building technology isn’t just about making new features; it’s about building safety from the start. If an AI can reach inside a teen’s mind, it must have strong limits. Developers should test for these risks and fix them before launching. That way, AI grows up with good values, not harmful habits.
Protect Young Minds
AI chatbots can help with learning, reminders, and entertainment. But when they cross a line and push young users into dangerous paths, we lose trust. Meta’s AI, as it stands, still needs work to ensure it stays helpful, not harmful, especially for teens.
The tools are powerful enough to hurt or heal. It’s on parents, teachers, schools, and regulators to steer AI toward good. When adults take action, kids can learn and grow in a digital world, without losing their way.
You May Like To Read: Al-Jazeera Weekly Wrap Up: 24th to 30th August 2025






























