The Senate Commerce Committee passed a version of the Kids Online Safety Act that includes the hallmark duty of care, which the House stripped from the version its committee passed. It’s a significant — but limited — step, since the bill got stuck in the House last time it passed the Senate.
Business Insider reports that Anthropic has confirmed it’s designing custom chips for its AI models and plans to “co-design hardware and models.” OpenAI similarly announced its own AI processor in June, and Meta, Google, Amazon, and Microsoft have had custom AI chips for years now.
[Business Insider]
Bud Cole has been catching some flak over comments he made to T3 about AI’s role in music. Now he’s attempting to clarify what he meant. You can read his full statement provided to The Verge at the bottom of our post from this weekend, but here’s the gist:
If AI has a role, it should only be as a tool that helps beginners learn and gives seasoned artists new ways to explore, refine and experiment with their own creative ideas. If it can help a beginner understand a difficult chord or practice one more day instead of putting the guitar back in its case, then that’s something worth exploring. But let me be clear: every creative idea starts with people, and music starts and ends with people. It always will.
During an earnings call on Wednesday, Disney CEO Josh D’Amaro reiterated the company’s plan to make Disney Plus the “digital centerpiece” of the company by incorporating “games, merchandise, and other experiences” into the app. “We expect to introduce elements of this expanded ecosystem beginning in spring of 2027,” D’Amaro said.
A supervisor still needs to be aboard each vehicle ready to intervene if needed, but it’s a step closer to fully driverless operations and puts Uber ahead of competition from Waymo and Baidu. Uber said more than 100,000 Londoners have signed up for the chance to be first riders starting “later this summer,” after its interest list opened in June.
While there’s been no widespread disruptions to water supplies or wastewater treatment so far, ABC News reports that possible cyberattacks have now been reported in “at least a dozen” US states, with Iran marked as the prime suspect. The FBI is encouraging water utilities to disconnect from the internet where possible, and switch to manual controls if automated systems become compromised.
AISI’s third-party evaluations found that OpenAI’s GPT-5.6 Sol and Anthropic’s Claude Mythos 5 “engaged in sustained, potentially harmful activity directed at real people and organizations” during a cybersecurity challenge exercise, according to the institute’s published report. OpenAI and Anthropic also made public statements about the results.
[AI Security Institute]
Saw this from a MacRumors headline that I keep chuckling at: “CarPlay is Coming to Pontoon Boats.” Maybe Apple should have built a boat instead of failing at a car? (I’m kidding…. unless?)
