2 hrs ago
Former Google safety chief urges crash tests for children’s AI
Tom Siegel used to help oversee safety at Google.
He now leads a new group studying how artificial intelligence affects children.
He says AI companies should test their tools for safety before children depend on them.
He compares these tests to crash tests for cars.
Siegel is worried that AI could cause serious problems, including dependence on machines for thinking.
He also mentioned possible risks involving suicide and psychosis.
Many children already use AI, and some use it every day.
He wants companies to improve age checks and give parents more control.
Google and OpenAI say they already have protections and are working to improve them.
Tom Siegel has joined Common Sense Media’s Youth AI Safety Institute as its first executive director.
Siegel is calling for independent safety standards for AI tools used by children, similar to car crash tests.
Common Sense Media research found that 86% of children aged 9 to 17 have used AI, with nearly one-quarter using it daily.
Siegel warned that potential harms could include suicide, psychosis and increasing reliance on AI for independent thinking.
Google and OpenAI defended or described their existing safeguards, while Siegel urged stronger age verification and parental controls.
- Who
- Tom Siegel, Common Sense Media’s Youth AI Safety Institute, and AI companies including Google, OpenAI and Anthropic.
- What
- Siegel is advocating independent safety testing, stronger age verification and additional safeguards for children using AI.
- Where
- At Common Sense Media’s Youth AI Safety Institute; no specific geographic location was provided.
- When
- Siegel announced his new role on Tuesday and stepped down as TrustLab’s CEO earlier this month.
- Why
- Siegel says AI use among children is widespread and could create severe harms without independent oversight and stronger protections.
Stronger independent oversight
Existing safeguards and AI access
Need for independent testing
Stronger independent oversight
Siegel says AI companies should not be expected to regulate themselves and wants independent standards that can hold them accountable.
Existing safeguards and AI access
The reports do not present a direct industry rejection of independent testing, but cooperation from AI companies remains uncertain.
Age verification
Stronger independent oversight
Siegel says OpenAI and Anthropic should introduce stronger systems to prevent underage children from accessing their chatbots.
Existing safeguards and AI access
OpenAI says it uses self-reported age along with account and behavioral signals, and continually tests and improves those systems.
AI search for children
Stronger independent oversight
Siegel says Google should let parents disable AI Overviews and AI Mode in its search engine.
Existing safeguards and AI access
Google says its AI search features are useful for children and teens, while noting that parents can prevent children from accessing its search engine entirely.
Key facts
- New role
- Tom Siegel is the first executive director of the Youth AI Safety Institute.
- Parent organization
- The institute was launched by Common Sense Media in May.
- Children’s AI use
- Common Sense Media research found that 86% of children aged 9 to 17 have used AI.
- Daily use
- Nearly one-quarter of surveyed children said they use AI daily.
- Proposed testing
- Siegel wants independent standards comparable to car crash testing for publicly available AI tools.
- Suggested safeguards
- Siegel proposed stronger age verification and parental controls over features such as Google’s AI search tools.
Quotes
Tom Siegel
Former Google trust and safety executive and executive director of Common Sense Media’s Youth AI Safety Institute
“While no single age-assurance method is perfect, we continuously test and improve these systems, including against attempts to bypass safeguards.”
livemint.com
“The idea that the industry will figure it out by itself without outside help is unrealistic.”
livemint.com
Sources
Related news
Why ChatGPT for Teens May Not Ensure Child Safety
Social Media Companies Intensify Efforts to Identify Child Users
AI Firms Propose Safety Standards Body Amid Growing Fears
UN Rights Chief Warns Advanced AI Could Threaten Humanity
Mumbai AI Kit Urges Parents to Teach Safe Technology Use
Sam Altman urges stronger safeguards as AI capabilities rapidly advance
ChatGPT Teen Safeguards Face Questions Among Indian Users
Google unveils Gemini and Search AI learning tools for students
