Associated Press
China’s National Development and Reform Commission blocked Meta Platforms from acquiring AI startup Manus, ordering all parties to unwind the deal over concerns about transferring advanced technology and AI talent to a foreign company. Manus, a Singapore-based AI firm with Chinese roots, develops autonomous agent systems capable of performing complex tasks. Analysts say the move signals Beijing’s willingness to restrict outbound tech assets, mirroring U.S. controls on technology deals with Chinese firms.
From "China Blocks Meta from Acquiring Startup Manus"
Associated Press (04/27/26) Chan Ho-Him; Kanis Leung; Kelvin Chan
View Full Article
Reuters
South Africa has withdrawn its first draft national AI policy after discovering that parts of its reference list contained fabricated, likely AI-generated, sources, undermining the document’s credibility. Communications Minister Solly Malatsi (pictured) said the inclusion of unverified citations highlighted the need for strong human oversight when using AI. The draft policy was framed as positioning the country as an African leader in AI innovation, outlining plans to establish new institutions and create incentives to encourage private-sector collaboration.
From "South Africa Withdraws AI Policy Due to Fake AI-Generated Sources"
Reuters (04/27/26) Nellie Peyton
View Full Article
Ars Technica
Researchers at Switzerland's École Polytechnique Fédérale de Lausanne have developed a control framework that allows robots to transfer learned skills across differently configured robots without the need for retraining. Unlike traditional or AI-driven approaches, the Kinematic Intelligence framework embeds each robot’s physical constraints—such as joint limits—directly into its control logic. By mapping robots into categories based on their mechanics, the software ensures tasks can be adapted across different designs.
From "New Control Software Avoids Jamming Robot Joints"
Ars Technica (04/26/26) Jacek Krywko
View Full Article
404 Media
In a study published online, researchers from Stanford University, Imperial College London in the U.K., and the Internet Archive said they found that by mid-2025, more than a third of websites created since 2022 were AI-generated or AI-assisted. Analyzing data from the Internet Archive, the team also found that AI content is making online writing less semantically diverse, while contrary to expectations, the researchers found no significant increase in false information or missing citations.
From "A Third of New Websites are AI-Generated: Study"
404 Media (04/27/26) Matthew Gault
View Full Article
Federal News Network
The U.S. National Institute of Standards and Technology (NIST) is launching a cybersecurity project to help critical infrastructure organizations improve visibility into their operational technology (OT) environments. Led by its National Cybersecurity Center of Excellence (NCCoE), the initiative is a response to widespread challenges in tracking and managing OT assets, especially in complex, legacy systems.
From "NIST Cyber Center to Launch OT ‘Visibility’ Project"
Federal News Network (04/23/26) Justin Doubleday
View Full Article
The Engineer (U.K.)
A 3-D bioelectronic device developed by Princeton University researchers combines living brain cells with embedded electronics. Built from a flexible mesh of microscopic wires and electrodes, the system supports the growth of tens of thousands of neurons into a dense three-dimensional network, allowing precise stimulation and long-term monitoring of neural activity. Over six months, scientists trained the evolving system to recognize patterns of electrical signals, successfully distinguishing spatial and temporal inputs.
From "Researchers Create a 3-D Bioelectronic Brain Cell Device"
The Engineer (U.K.) (04/24/26)
View Full Article
