AILegal

Authors File Class Action Against Salesforce Over AI Training Data

Salesforce faces a proposed class action lawsuit from authors claiming the company used pirated books to train its AI software. The case adds to mounting legal challenges against tech firms over AI training practices.

Legal Challenge Over AI Training Materials

Cloud computing giant Salesforce is facing a proposed class action lawsuit from two authors who allege the company used thousands of copyrighted books without permission to train its artificial intelligence systems, according to reports filed this week.

AIEthics

Study Reveals Widespread Inability to Detect Racial Bias in AI Training Data

Researchers found that artificial intelligence systems can develop racial bias when trained on unrepresentative data, but most users fail to recognize these imbalances. The study reveals that people typically only notice bias when AI systems demonstrate skewed performance in classifying emotions across different racial groups.

AI Systems Learn Racial Bias Through Training Data

According to a recent study published in Media Psychology, most users cannot identify racial bias in artificial intelligence training data, even when it’s clearly presented to them. The research indicates that AI systems can develop skewed perceptions, such as classifying white people as happier than individuals from other racial backgrounds, due to imbalanced training datasets.

AITechnology

Tech Giants Forge New Ethernet Alliance to Power Next-Generation AI Infrastructure

A coalition of technology leaders including Meta, Microsoft, and NVIDIA has launched the Ethernet for Scale-Up Networking initiative to advance AI infrastructure capabilities. The collaboration aims to establish open standards for handling massive AI workloads across distributed systems. Meta engineers simultaneously revealed significant advancements in their data center networking architecture.

Industry Leaders Unite on Ethernet Standards for AI Scaling

Major technology companies have formed a new alliance to develop Ethernet standards specifically designed for scaling artificial intelligence infrastructure, according to reports from the Open Compute Project. The Ethernet for Scale-Up Networking (ESUN) initiative brings together Meta Platforms, AMD, Arista, ARM, Broadcom, Cisco, HPE Networking, Marvell, Microsoft, NVIDIA, OpenAI and Oracle to address growing demands in AI system networking.

AITrade

AI-Powered Customs Solutions Help Importers Navigate Tariff Complexity

U.S. businesses are increasingly turning to AI-driven customs brokerage services to handle complex tariff regulations. These tools reportedly help importers optimize supply chains and reduce duty expenses through automated analysis and real-time recommendations.

Rising Demand for AI in Trade Compliance

According to reports, U.S. importers are adopting specialized artificial intelligence tools to manage increasing tariff costs and regulatory complexity. Industry sources indicate that since the detailed implementation of sweeping import tariffs in August, businesses have sought technological solutions to limit financial impacts. These tariff increases have elevated costs for imported goods and materials, prompting companies to explore automated compliance and optimization platforms.