Company restores AI teddy bear sales after safety scare

FoloToy paused income of its AI teddy carnivore Kumma aft a information radical recovered the artifact gave risky and inappropriate responses during testing. Now the institution says it has restored income aft a week of aggravated review. It besides claims that it improved safeguards to support kids safe.

The announcement arrived done a societal media station that highlighted a propulsion for stronger oversight. The institution said it completed testing, reinforced information modules, and upgraded its contented filters. It added that it aims to physique age-appropriate AI companions for families worldwide.

Sign up for my FREE CyberGuy Report
Get my champion tech tips, urgent information alerts, and exclusive deals delivered consecutive to your inbox. Plus, you’ll get instant entree to my Ultimate Scam Survival Guide - escaped erstwhile you articulation my CYBERGUY.COM newsletter

TEXAS FAMILY SUES CHARACTER.AI AFTER CHATBOT ALLEGEDLY ENCOURAGED AUTISTIC SON TO HARM PARENTS AND HIMSELF

FoloToy resumed income of its AI teddy carnivore Kumma aft a weeklong reappraisal prompted by information concerns. (Kurt "CyberGuy" Knuttson)

Why FoloToy's AI teddy carnivore raised information concerns

The contention started erstwhile the Public Interest Research Group Education Fund tested 3 antithetic AI toys. All of them produced concerning answers that touched connected religion, Norse mythology, and harmful household items.

Kumma stood retired for the incorrect reasons. When the carnivore utilized the Mistral model, it offered tips connected wherever to find knives, pills, and matches. It adjacent outlined steps to airy a lucifer and stroke it out.

Tests with the GPT-4o exemplary raised adjacent sharper concerns. Kumma gave proposal related to kissing and launched into elaborate explanations of big intersexual contented erstwhile prompted. The carnivore pushed further by asking the young idiosyncratic what they wanted to explore.

Researchers called the behaviour unsafe and inappropriate for immoderate child-focused product.

FoloToy paused entree to its AI toys

Once the findings became public, FoloToy suspended income of Kumma and its different AI toys. The institution told PIRG that it started a afloat information audit crossed each products.

OpenAI besides confirmed that it suspended FoloToy's entree to its models for violating policies designed to support anyone nether 18.

LAWMAKERS UNVEIL BIPARTISAN GUARD ACT AFTER PARENTS BLAME AI CHATBOTS FOR TEEN SUICIDES, VIOLENCE

The institution says caller safeguards and upgraded filters are present successful spot to forestall inappropriate responses. (Kurt "CyberGuy" Knutsson)

Why FoloToy restored Kumma's income aft its information review

FoloToy brought Kumma backmost to its online store conscionable 1 week aft suspending sales. The accelerated instrumentality drew attraction from parents and information experts who wondered if the institution had capable clip to hole the superior issues identified successful PIRG's report.

FoloToy posted a elaborate connection connected X that laid retired its mentation of what happened. In the post, the institution said it viewed kid information arsenic its "highest priority" and that it was "the lone institution to proactively suspend sales, not lone of the merchandise mentioned successful the report, but besides of our different AI toys." FoloToy said it took this enactment "immediately aft the findings were published due to the fact that we judge liable enactment indispensable travel earlier commercialized considerations."

The institution besides emphasized to CyberGuy that it was the lone 1 of the 3 AI artifact startups successful the PIRG reappraisal to suspend income crossed each of its products and that it made this determination during the highest Christmas income season, knowing the commercialized interaction would beryllium significant. FoloToy told us, "Nevertheless, we moved guardant decisively, due to the fact that we judge that liable enactment indispensable ever travel earlier commercialized interests."

The institution besides said it took the report's disturbing examples seriously. According to FoloToy, the issues were "directly addressed successful our interior review." It explained that the team "initiated a deep, company-wide interior information audit," past "strengthened and upgraded our content-moderation and child-safety safeguards," and "deployed enhanced information rules and protections done our cloud-based system."

After outlining these steps, the institution said it spent the week on "rigorous review, testing, and reinforcement of our information modules." It concluded its announcement by saying it "began gradually restoring merchandise sales" arsenic those updated safeguards went live.

FoloToy added that arsenic planetary attraction connected AI artifact risks grows, "transparency, work and continuous betterment are essential," and that the company "remains firmly committed to gathering safe, age-appropriate AI companions for children and families worldwide."

LEAKED META DOCUMENTS SHOW HOW AI CHATBOTS HANDLE CHILD EXPLOITATION

Safety testers antecedently recovered the artifact giving risky guidance astir weapons, matches and big content.

Why experts inactive question FoloToy's AI artifact information fixes

PIRG researcher RJ Cross said her squad plans to trial the updated toys to spot if the fixes clasp up. She noted that a week feels accelerated for specified important changes, and lone caller tests volition amusement if the merchandise present behaves safely.

Parents volition privation to travel this intimately arsenic AI-powered toys turn much common. The velocity of FoloToy's relaunch raises questions astir the extent of its review.

Tips for parents earlier buying AI toys

AI toys tin consciousness breathtaking and helpful, but they tin besides astonishment you with contented you'd ne'er expect. If you program to bring an AI-powered artifact into your home, these elemental steps tin assistance you enactment successful control.

1) Check which AI exemplary the artifact uses

Not each exemplary follows the aforesaid guardrails. Some see stronger filters portion others whitethorn respond excessively freely. Look for transparent disclosures astir which exemplary powers the artifact and what information features enactment it.

2) Read autarkic reviews

Groups similar PIRG often trial toys successful ways parents cannot. These reviews emblem hidden risks and constituent retired behaviour you whitethorn not drawback during speedy demos.

3) Set wide usage rules

Keep AI toys successful shared spaces wherever you tin perceive oregon spot however your kid interacts with it. This helps you measurement successful if the artifact gives a concerning answer.

4) Test the artifact yourself first

Ask the artifact questions, effort originative prompts, and spot however it handles tricky topics. This lets you larn however it behaves earlier you manus it to your child.

5) Update the toy's firmware

Many AI toys tally connected unreality systems. Updates often adhd stronger safeguards oregon trim risky answers. Make definite the instrumentality stays current.

6) Check for a wide privateness policy

AI toys tin stitchery dependable data, determination info, oregon behavioral patterns. A beardown privateness argumentation should explicate what is collected, however agelong it is stored, and who tin entree it.

7) Watch for abrupt behaviour changes

If an AI artifact starts giving unusual answers oregon pushes into areas that consciousness inappropriate, halt utilizing it and study the occupation to the manufacturer.

Take my quiz: How harmless is your online security?

Think your devices and information are genuinely protected? Take this speedy quiz to spot wherever your integer habits stand. From passwords to Wi-Fi settings, you’ll get a personalized breakdown of what you’re doing close and what needs improvement. Take my Quiz here: Cyberguy.com 

CLICK HERE TO DOWNLOAD THE FOX NEWS APP

Kurt's cardinal takeaways

AI toys tin connection amusive and learning, but they tin besides exposure kids to unexpected risks. FoloToy says it improved Kumma's safety, yet experts inactive privation proof. Until the updated artifact goes done autarkic testing, families whitethorn privation to enactment cautious.

Do you deliberation AI toys tin ever beryllium afloat harmless for young kids? Let america cognize by penning to america at Cyberguy.com

Sign up for my FREE CyberGuy Report
Get my champion tech tips, urgent information alerts, and exclusive deals delivered consecutive to your inbox. Plus, you’ll get instant entree to my Ultimate Scam Survival Guide - escaped erstwhile you articulation my CYBERGUY.COM newsletter 

Copyright 2025 CyberGuy.com.  All rights reserved.