Stay up to date with the latest developments in artificial intelligence.
Sondera announced that its research on compiling natural language policies into formally verified rules for AI agents has been accepted at ICML 2026 and FLoC 2026, with a related tool demonstration at Black Hat Arsenal.
Inductive Bio has joined Anthropic's life sciences ecosystem and launched a connector that integrates its ADMET prediction models into Claude, allowing scientists to analyze drug properties directly within the AI assistant.
Basecamp Research has made its EDEN antibiotic and vaccine design models available through Anthropic's Claude Science, allowing researchers to generate and prioritize therapeutic candidates using a conversational interface.
Sprocket Security has introduced Apex, an AI penetration testing agent designed for continuous, human-supervised web application testing. The system identifies exploitable vulnerabilities using customer context and historical data while operating under a zero data retention policy.
Siteimprove has announced the Siteimprove.ai Model Context Protocol (MCP) Server, connecting its Accessibility Agent directly into AI and design environments such as Anthropic Claude, Lovable, VS Code, and Figma to improve digital accessibility during content creation.
Kodex has unveiled a new infrastructure to manage government and law enforcement requests for user data, featuring automated processing, direct system intake, and agent workflows.
UBTECH has launched the UWORLD U1 Series, which it calls the world’s first mass-produced full-size ultra-bionic humanoid robot lineup, featuring models designed for both consumer and commercial use.
SciBase has submitted a 510(k) premarket notification to the U.S. FDA to expand the use of its AI device Nevisense to include non-melanoma skin cancers, aiming to broaden its clinical application beyond melanoma.
Verkada has announced a collaboration with NVIDIA to accelerate the development and deployment of its physical AI platform, which operates across more than 2.4 million devices globally. NVIDIA also joins as a new investor in Verkada.
NetRise has announced a partner-led managed software supply chain risk management offering for the federal market, combining binary analysis with supply chain provenance to help agencies operationalize software risk assessment.
National Grid Ventures will invest $1.75 billion for a 35% stake in Joulent to develop large-scale power and electrical infrastructure for US data centers, including a 2.67 GW facility in West Texas supplying a Microsoft data center.
Netzilo has expanded its AI Detection and Response platform to Amazon Bedrock AgentCore and other AI agent frameworks, introducing Bring Your Own Governance to enterprise AI agents for consistent observation, detection, and response across environments.
PreOncology has introduced a new oncologist-led medical specialty focused on preventing and detecting cancer before stage I, combining genomics, AI risk modeling, and personalized surveillance through its Stage Alpha Operating System.
DriveNets has introduced two new AI networking platforms, the 2600SL and 2601S, built on Broadcom’s Tomahawk 6 ASIC, offering up to 102.4Tbps capacity and designed for large-scale AI infrastructure deployments.
LatusPet, a University of Oxford biotech startup, has published a peer-reviewed study demonstrating that its SINO platform can detect multiple diseases in dogs from a single blood sample with over 90% accuracy.
Caris Life Sciences has launched Caris Detect, a blood test that uses whole genome and transcriptome sequencing with artificial intelligence to identify early cancer signals across multiple cancer types.
Cloud Capital has launched a Core Joint Venture Strategy with Realty Income and a global institutional investor, seeded with over $6 billion of stabilized hyperscale data center assets in the United States, with plans to expand into Europe.
Anthropic has reinstated access to its Claude Fable 5 model following the lifting of U.S. export controls that had suspended its availability earlier in June. The company introduced new safeguards and is working with industry and government partners on a standardized framework for assessing AI jailbreaks.
Anthropic has launched Claude Sonnet 5, a new model that improves on Sonnet 4.6 with stronger reasoning, coding, and tool use abilities while maintaining lower costs and enhanced safety features. The model is now available across all plans with introductory pricing through August 31, 2026.
Taxwire has raised $25 million in combined Seed and Series A funding to scale its AI-powered sales tax automation platform, which helps businesses manage compliance across 22,000 U.S. jurisdictions and over 100 countries.
Daily report covering major AI developments and industry news, with both top stories and complete market updates
