OpenAI Begins Rolling Out Age Prediction on ChatGPT to Expand Teen Safety Protections

OpenAI is using behavioral signals—not just self-reported age—to determine when teen safety protections should apply on ChatGPT.

Google AI Overviews Flagged for Health Misinformation After Investigation Finds Medical Errors

What happens when AI-generated health advice appears first—and gets it wrong?

OpenAI Hardens ChatGPT Atlas Against Prompt Injection With Automated Red Teaming

As AI agents take on more real-world tasks, OpenAI is racing to secure ChatGPT Atlas against prompt injection attacks that can quietly steer agents off course without a user ever realizing it.

Google Explains How Chrome Secures Agentic AI Features With Human Oversight & Guardrails

As browsers begin taking actions on users’ behalf, Google is outlining how Chrome’s agentic AI features are being designed to prioritize security, transparency, and human control.

Amazon Ring Launches Facial Recognition for Doorbells, Raising Security and Privacy Questions

Amazon is bringing facial recognition directly to the front door, as Ring rolls out its Familiar Faces feature to identify who’s approaching a home — and reigniting debates around privacy, security, and biometric data in everyday spaces.

How BrowseSafe Detects Prompt Injection Threats in AI Browser Agents

AI browser agents now navigate the same cluttered, unpredictable webpages users do—making prompt-injection detection essential for protecting real online actions.

OpenAI Drops Mixpanel After Security Incident Exposes Limited User Metadata

OpenAI is notifying API customers about a security incident inside Mixpanel’s systems that exposed limited account metadata—but did not compromise any chat content, API keys, credentials, or payment information.

OpenAI Unveils Aardvark, a GPT-5 Agent for Proactive Code Security

OpenAI has launched Aardvark, a GPT-5-powered autonomous security agent designed to detect and remediate software vulnerabilities across modern codebases.

Meta Expands Parental Controls for Teen AI Use Across Its Platforms

Meta is expanding its commitment to AI safety for teens, introducing new parental supervision tools that allow families to monitor, manage, and guide how young users engage with AI characters across the company’s platforms.

Google Adds AI-Powered Ransomware Protection to Drive for Desktop

Ransomware accounted for 21% of intrusions last year, with the average incident costing more than $5 million — a risk Google now aims to counter with AI-powered defenses in Drive for desktop.

DeepMind Expands Frontier Safety Framework With New AI Risk Domains

DeepMind has released the third iteration of its Frontier Safety Framework, adding new domains such as harmful manipulation and expanding misalignment protocols to strengthen governance of advanced AI models.

OpenAI Balances Teen Safety, Privacy, and Age Prediction in AI

OpenAI is prioritizing teen safety by introducing age prediction systems and parental controls, while reaffirming commitments to privacy and user freedom.

California and FTC Target AI Chatbots with New Safeguards

California moves toward first-in-the-nation AI companion chatbot law as the FTC launches a federal inquiry into children’s safety.

Google Gemini Rated ‘High Risk’ for Kids and Teens by Common Sense Media

A new safety review warns that Google’s Gemini AI exposes children and teens to inappropriate content, despite added safeguards.

Anthropic Backs California’s AI Safety Bill SB 53, Breaking with Tech Opposition

Anthropic’s endorsement of SB 53 marks a rare win for state AI regulation, as the bill advances toward a final vote.

OpenAI adds parental controls and expert guidance to ChatGPT

OpenAI is introducing parental controls for ChatGPT, alongside new safeguards for sensitive conversations and expanded expert guidance on mental health.

OpenAI and Anthropic Share Models for Joint AI Safety Testing

OpenAI and Anthropic briefly shared access to their AI models for joint safety testing — a rare collaboration to expose blind spots and set new safety standards.

Anthropic Pilots Claude for Chrome with Safety Controls

Anthropic is testing a Chrome extension that lets Claude act inside the browser, while confronting security risks like prompt injection attacks.

Character.AI Weighs Sale or New Funding Amid Legal and Cost Pressures

The AI chatbot maker is considering a billion-dollar valuation deal while lawsuits and high operating costs raise questions about its future.

Claude Models Can Now End Conversations in Extreme Cases, Says Anthropic

Anthropic has introduced a new safeguard in Claude 4 and 4.1 models, allowing them to terminate conversations under rare and extreme conditions, marking a shift in how AI handles harmful dialogue.

California Lawmakers Advance Bill to Regulate Harmful AI Chatbot Practices

A bipartisan privacy committee has approved Senate Bill 243, aimed at curbing addictive and unsafe design features in AI companion chatbots.

OpenAI Boosts Internal Security After Suspected IP Theft

OpenAI adopts stricter controls on data, staff access, and office protocols following concerns about espionage and leaks.

Denmark Moves to Let Citizens Copyright Their Own Faces, Voices

Denmark is preparing legislation to let individuals copyright their face, voice, and body—a direct response to deepfakes and generative AI misuse.

U.S. Says AI Firm DeepSeek Aids Chinese Military, Evades Chip Controls

Reddit Weighs Adopting Sam Altman’s Iris-Scanning Orb for User Verification