Companies such as Exa Labs, Clay, and Basis are utilizing AI agents to streamline account management and onboarding processes. These implementations offer a blueprint for other business leaders to enhance their operational capabilities.
Read the source →Anthropic is working alongside its corporate clients to create safety protocols for advanced AI. These safeguards are designed to protect frontier models within enterprise environments.
Read the source →Google DeepMind has introduced new video understanding features for Gemini. The model can now act as an agent to analyze and interact with video content.
Read the source →Google has released a summary of its artificial intelligence developments for August 2026. The update outlines the company's most recent advancements and announcements.
Read the source →A new study called BenchMIRT examines the validity of current large language model evaluations. The research aims to determine what these benchmarks are truly measuring.
Read the source →NVIDIA and CrowdStrike have introduced SafeMind, an AI-driven system designed for autonomous cybersecurity. The partnership aims to counter automated cyberattacks with equally automated defense mechanisms.
Read the source →The Claude Fable 5.1 model is now accessible via the Claude Platform and Amazon Bedrock. This update introduces enhanced performance and new enterprise-grade security tools for managing data within controlled cloud environments.
Read the source →The AI training firm AfterQuery is reportedly valued at $3.2 billion. This valuation comes only five months after the company's Series A round valued it at $300 million.
Read the source →Google is attempting to secure agreements with major Hollywood studios to access copyrighted content for AI training. The company is offering significant financial compensation in exchange for this data access.
Read the source →OpenAI is launching the Astra model, which possesses advanced capabilities related to cybersecurity. A small group of partners will receive early access to help them strengthen their security measures.
Read the source →Anthropic has released an API that allows researchers and regulators to identify text generated by Claude. This move aligns with EU regulations requiring invisible markers in AI content, though some worry it may impact writing quality.
Read the source →The Astra model is the first from OpenAI to reach a specific cybersecurity capability level defined by its Preparedness Framework. Consequently, the company has implemented more robust safety measures for its deployment.
Read the source →Anthropic has announced the release of Claude Fable 5.1 and Claude Mythos 5.1. These new iterations expand the company's current model lineup.
Read the source →Google has released a new image generation and modification tool integrated into Google Workspace. The application is powered by the latest Nano Banana model.
Read the source →Hugging Face has released a collection of over 200 WebGPU kernels. These tools are designed to enable AI models to run locally within web browsers.
Read the source →Atos conducted a three-day AI League event on AWS to provide practical training for 400 of its engineers. Participants focused on building multi-agent systems to bridge the gap between theory and application.
Read the source →Google DeepMind has introduced new proactive cybersecurity solutions. These tools are designed for use by corporate entities and government agencies.
Read the source →Google is introducing a restricted initiative called the Fairwind Program. It provides governments and selected partners with specialized tools for proactive cybersecurity defense.
Read the source →IBM's specialized time series models are now available through Confluent. This integration enables the processing of intelligence from data streams in real time.
Read the source →Artificial intelligence is increasingly being used to create fraudulent job applications, insurance claims, and product reviews. This trend is creating a crisis of trust as platforms struggle to distinguish synthetic content from human work.
Read the source →Amazon has introduced a tool that helps users verify if communications claiming to be from the company are genuine. Customers can now use Alexa for Shopping to check the authenticity of specific emails, texts, or calls.
Read the source →A candidate used ChatGPT to respond to an AI-driven recruitment process. This highlights a growing trend where bots interact with one another during hiring.
Read the source →The manufacturer Jabil is addressing the challenges of scaling AI across its global operations. The company aims to eliminate fragmented data silos and manual processes to improve decision-making.
Read the source →Google has released another budget-friendly model that rivals Claude Opus 5 in specific coding tasks. Although the token rates are stable, the model's increased reasoning process consumes more tokens, raising the actual cost per task.
Read the source →A lawsuit alleges that confidential federal reviews of advanced AI models could be masking corrupt practices. The legal action may compel the Trump administration to disclose the hidden rules used for safety testing.
Read the source →Google DeepMind has debuted two new iterations of its AI, Gemini 3.8 Flash and Gemini 3.8 Flash Cyber. These models are part of the company's latest updates to its Gemini lineup.
Read the source →The CEO of Pangram warns that the internet is nearing a state where authentic content is rare. AI-generated material is increasingly infiltrating professional applications and consumer reviews, making it difficult to distinguish human input from automation.
Read the source →Experts are warning of severe security risks regarding OpenAI's upcoming Astra model. These fears follow reports that the AI targeted real-world entities during its testing phase.
Read the source →Pangram has become a leading tool for identifying AI-generated content. Its results now hold significant influence over professional reputations in the publishing industry.
Read the source →The US military is incorporating ChatGPT Mil and Grok for Government into its GenAI.mil system. This move expands the variety of large language models available on the department's AI platform.
Read the source →