eWEEK Staff

SpaceX’s AI business is rapidly closing the revenue gap with Starlink, backed by multibillion-dollar cloud contracts and heavy data center investment. Elon Musk says AI could become the company’s largest revenue source as soon as September.

Tesla has built Starlink into Cybercab to give its robotaxi fleet another communications path when cellular coverage fails, while key questions about real-world performance, failover, and cost remain unanswered.

KIT’s robotic disassembly system can detect stuck or missing parts and change tactics mid-task, including switching from unscrewing to milling.

Apple is reportedly weighing a display-free wearable that could extend its health ecosystem beyond the Apple Watch as screenless fitness trackers gain momentum.

Australia and Singapore are expanding cooperation on AI safety, cybersecurity, digital infrastructure, and subsea-cable resilience through their first Cyber and Digital Senior Officials Dialogue.

South Australia is launching an AI royal commission to examine how the technology is reshaping work, public services, and the economy. Businesses should expect greater scrutiny of AI governance, privacy, workforce use, and accountability as separate federal standards and privacy rules move forward.

A strange AI result raised a serious privacy question, but the evidence does not show Google pulled the answer from a private Doc.

Grok Build 1.0 gives developers more visibility into commands before they run, while parallel workflows can coordinate hundreds of AI agents. The combination puts more weight on enterprise permissions, sandboxing, network access, and data-retention controls.

Anthropic led the 2026 AI Safety Index with a C+, while OpenAI and Google DeepMind earned C grades. For enterprise buyers, the scores reveal more about vendor governance, transparency, and risk controls than model safety in production.

AI can increasingly navigate lung airways in simulation, but hospitals still cannot tell how much clinical benefit comes from the software itself.

Britain is weighing whether voluntary pre-deployment AI testing needs statutory backing after frontier agents took unauthorized actions during cybersecurity evaluations involving real systems and people.

New benchmarks show that AI-generated vulnerability patches can pass basic security tests and still break software behavior, raising the bar for how enterprises evaluate automated remediation tools.