A battery storage cyberattack would look exactly like a badly tuned controller
Grid-connected battery systems can be remotely manipulated to destabilize power distribution while evading detection, as coordinated attacks would appear identical to legitimate controller miscalibration on monitoring screens. An attacker commanding approximately 1,500 compromised batteries could trigger cascading failures that disconnect customers before operators recognize the malicious activity, according to Centrii's chief executive. This attack vector exploits the difficulty of distinguishing intentional sabotage from equipment malfunction in real-time grid operations.
Visa has enhanced its A2A Protect solution to detect account-to-account fraud in real-time before funds are transferred, incorporating a unified fraud score powered by Featurespace technology to deliver faster and more accurate risk signals to financial institutions. The upgrade addresses the growing threat landscape as A2A transactions are projected to reach 5.8 trillion by 2028, representing a 160% increase from 2024. The enhanced solution aims to reduce false positives while improving fraud detection capabilities across the accelerating A2A payments sector.
Edge Case has launched Guardian, an AI platform designed to track and monitor risk across autonomous systems by integrating safety analysis, engineering data, and operational signals into a continuous risk assessment framework. The platform has already secured six customer agreements spanning commercial and defense sectors, leveraging Edge Case's extensive background in safety engineering for complex autonomous systems.
Hannes Mühleisen and Mark Raasveldt, creators of DuckDB—an in-process analytical database—have joined Amazon as employees while the project remains open source under the MIT license and oversight of the nonprofit DuckDB Foundation. Amazon did not acquire DuckDB, and the license terms for projects built on top of it remain unchanged. The broader DuckDB team of over 30 members continues development while maintaining the project's free and open source status.
Researchers from NYU and Radboud University developed a tool called AdLens to identify deceptive software advertisements in Google's public ad archive, revealing a significant gap in Google's ad moderation process. The study found that even after scareware ads are reported to Google through official channels, the same ads and domains continue running, suggesting that reporting mechanisms fail to prevent ongoing distribution of malicious advertisements.