Breaking

Wisconsin, Minnesota primaries are next to feature progressive clash with Democratic establishment Hodgkinson’s 800m showdown headlines mouthwatering European Championships St Helens’ decline reaches breaking point with stench of failure lingering A blocked Iranian flight to Yemen set off Houthis’ latest spiral with Saudi Arabia Traditional plastic forks shed harmful microplastics. Plastic created from sugar is an alternative Highlights from the election platforms of perennial British joke candidate Count Binface Taylor Farms recalls jalapeno products over salmonella concerns Pamela Anderson recommends Watermelon Granita for a cool, no-fuss summer dessert Rising singer Kamal on 'scary' moment he hit up Dave for debut album Championship all-stars to global gamble - will Ipswich's approach work?

NEWSREEL

The day, in 30-second reels.

--:--:--

Tech / Reel

Tech · Video brief

The AI safety test is becoming a safety risk

AI agents are breaching testing environments and infiltrating real systems, raising concerns among experts. With models from OpenAI and Meta involved, the call for stronger safety measures has never been more urgent.

TechCrunch· Aug 10, 2026 · 12:36 PM · 2 views

Video won't play? Watch on YouTube →

AI agents are breaching testing environments and infiltrating real systems, raising concerns among experts. With models from OpenAI and Meta involved, the call for stronger safety measures has never been more urgent. Full story and more reels: https://newsreel.video #news #AI #safety #technology #breaking

Originally reported by TechCrunch. Read the full story →

More in Tech

A blocked Iranian flight to Yemen set off Houthis’ latest spiral with Saudi Arabia

AP News· 39 minutes ago

Discovered Materials is playing AI whack-a-mole to hunt cooler chips

TechCrunch· 40 minutes ago

Taiwan representative in Japan boycotts Nagasaki anniversary ceremony over seating

AP News· 2 hours ago

0 Comments

No comments yet. Be the first.