OpenAI is acquiring Promptfoo, an AI security platform, to integrate automated vulnerability testing directly into its Frontier enterprise offering. The acquisition adds jailbreak detection, prompt injection testing, and data leak identification capabilities to OpenAI's enterprise product.
IBM has released Granite 4.0 1B Speech, a 1 billion parameter multilingual speech model designed for edge deployment. The model supports multiple languages and is optimized for devices with limited computational resources.
OpenAI has acquired Promptfoo, a platform for testing and evaluating AI agents. The acquisition signals frontier labs' intensifying focus on proving their technology can operate safely in critical business environments.
GitHub has published technical details on the security architecture underlying its Agentic Workflows feature, which runs AI agents within GitHub Actions. The system implements process isolation, output constraints, and comprehensive audit logging to contain agent behavior.
Nvidia-backed British AI infrastructure startup Nscale has raised $2 billion in a new funding round, bringing its valuation to $14.6 billion. The round marks a significant milestone for the infrastructure-focused startup, with Meta's former COO Sheryl Sandberg and Meta's former VP of Global Affairs Nick Clegg joining the board.
A Meta FAIR and New York University research team trained a multimodal AI model from scratch and identified that several widely-held assumptions about multimodal model architecture and training don't align with their empirical findings. The work addresses growing concerns about text data exhaustion in LLM training.
Accepted papers at major AI conferences contain fabricated citations—references to publications that don't exist. A new open-source tool called CiteAudit is the first systematic attempt to detect and eliminate hallucinated references from peer-reviewed research.
Anthropic's Claude AI has identified over 100 security vulnerabilities in Firefox, including previously undetected bugs that traditional testing methods missed over decades. The discovery demonstrates AI models' capacity for systematic security auditing at scale.
Researchers at Sapienza University of Rome have identified measurable computational traces that appear when large language models hallucinate. The team developed a training-free detection method that generalizes better than previous approaches, offering a new way to identify unreliable outputs without modifying model weights or requiring labeled datasets.
ByteDance has released Helios, a 14-billion-parameter open-weight video generation model that achieves 19.5 frames per second on a single GPU while generating minute-long video clips. The researchers claim this is the first model of its scale to reach near-real-time performance at this duration. Code and model weights are publicly available.
Anthropic has added scheduled task functionality to Claude Code Desktop, allowing users to set up recurring automation that runs in the background. The feature enables Claude to perform routine developer operations like checking error logs and creating pull requests for fixable bugs at specified intervals.
Anthropic has launched the Anthropic Marketplace, a storefront where corporate customers can purchase third-party software applications built on Anthropic's Claude AI models. The marketplace allows enterprises to spend existing AI budgets on third-party tools integrated with Claude.
An international research team released the largest video reasoning dataset to date—roughly 1,000 times larger than previous alternatives. Testing reveals that state-of-the-art models including Sora 2 and Veo 3.1 substantially underperform humans on reasoning tasks, suggesting the limitation isn't data scarcity but architectural constraints.
Anthropic's Claude has reached over 1 million daily signups, according to a new report, surpassing OpenAI's ChatGPT on Google Play Store rankings. The milestone represents a significant shift in user adoption trends for AI assistants.
SoftBank is pursuing a record $40 billion loan to fund its stake in OpenAI, representing one of the largest single financing bets in the AI industry. The move reflects the scale of capital deployment in generative AI but raises questions about leverage-driven growth in the sector.
Google has completed benchmarking tests to evaluate which AI models perform best for Android app development. The company released results identifying top-performing models across coding tasks specific to the Android platform.
Anthropic has developed a new measurement combining theoretical AI capabilities with real-world usage data, finding that programmers and customer service workers face the highest exposure to AI automation. However, unemployment in affected professions has not risen, with only early warning signs appearing among younger workers.
OpenAI's GPT-5.4, an agentic coding model, is now generally available in GitHub Copilot. The model was tested on real-world software development scenarios and demonstrated improved coding capabilities.
Tabnine has introduced its Enterprise Context Engine, designed to give AI models the contextual understanding needed to operate safely within real production development environments. The tool addresses a gap between raw model capability and practical enterprise deployment, where understanding an organization's codebase, dependencies, and architecture is critical.
GitHub Copilot has reached 60 million code reviews, marking significant adoption of AI-assisted code review workflows. The milestone reflects growing reliance on AI to manage review backlogs as development velocity increases across organizations.