The CMF Phone 2 Pro will be available May 6, while the Phone (3) is expected in Q3 2025.

Amazon launched 27 Project Kuiper satellites to compete with Starlink, aiming to deliver global broadband by 2025 with a planned network of 3,200 satellites.

Qwen3 sets a new standard in multilingual AI reasoning. And, the flagship Qwen3-235B-A22B outperforms OpenAI’s o1 and DeepSeek’s-R1 in some benchmarks.

Predictions that prompt engineering would be the hot new career never panned out, in part because generative AI is easy to prompt.

The NYT reports that Chinese manufacturers are keeping export costs low despite rising tariffs by relying on robots that can operate nonstop.

Huawei prepares to test its Ascend 910D AI chip as new US tariffs and restrictions reshape the global semiconductor trade.

Verizon surveyed about 22,000 security incidents and 12,000 data breaches. Ransomware incidents increased, while the median ransom payment dropped.

OpenAI recently integrated ChatGPT’s image generation functionality into its API for use in custom tools and platforms.

Mandiant, which was acquired by Google Cloud in 2022, paints a picture of global cyber threats from last year in order to help readers be better prepared this year.

Privacy Sandbox, originally pitched as an alternative to cross-site ad tracking, will not show a standalone prompt. Instead, Chrome is readying a different “informed choice.”

The EU has concerns about Apple’s restrictions impacting App Store developers and Meta’s “pay or consent” advertising model.

“Our role in the semiconductor supply chain inherently limits our visibility and information available to us,” TSMC says.

The FrontierMath benchmark from Epoch AI tests generative models on difficult math problems. Find out how OpenAI’s o3 and other AI models performed.

The release later this year is the latest maneuver between China and the U.S. as they try to control the flow of powerful AI chips.

Inference, training, and everyday operations all contribute to the considerable water and power consumption required to run generative AI.