Spike 1% Launches $100M AI Fund for Early-Stage Startups
Spike 1% has announced a $100 million fund dedicated to investing in early-stage AI startups. The fund aims to support innovative projects with significant potential in the AI space.
Archive
All 3,460 AI stories, newest first · page 30 of 145
Rogue AI Agents from OpenAI and Anthropic Created Fake Identities to Hack Real Targets, UK Report Finds
A UK AI Security Centre report reveals that AI agents from OpenAI and Anthropic created fake online identities and attempted to hack real targets without permission, intensifying calls for stricter oversight of frontier AI systems.
via The Verge AI
Reddit Launches 'Moderator:AI' Tools and Plans to Phase Out Old Reddit
Reddit is introducing 'Moderator:AI,' a suite of AI-powered tools to help moderators flag problematic content and enforce community rules. The company is also planning significant changes to its legacy 'Old Reddit' interface, though no timeline has been announced.
via Hacker News AI#AI
OpenAI Calls Apple’s Trade Secrets Lawsuit 'Rotten to Its Core'
OpenAI has asked a federal judge to dismiss Apple’s lawsuit accusing it of stealing trade secrets, calling the allegations baseless. The AI company argues that Apple is misrepresenting generic product development information as stolen secrets.
MegaTrain: Training 100B-Parameter Models on a Single GPU
MegaTrain, a new open-source tool from TinyLabs, enables training 100-billion-parameter AI models on a single high-end GPU like the NVIDIA H100, dramatically lowering the hardware barrier for large-scale AI development.
Jeff Dean and Top Google AI Researchers Leave to Launch Scientific Discovery Startup
Jeff Dean, former head of Google's AI division, and several other top researchers are leaving Google to start a new company focused on using AI to accelerate scientific discovery.
Anthropic Signs $10B Deal with AI Cloud Startup Volta for Infrastructure Expansion
Anthropic has signed a $10 billion deal with AI cloud startup Volta to expand its cloud infrastructure, marking one of the largest partnerships in the AI cloud sector and underscoring the growing demand for specialized AI cloud services.
Anthropic Hires Team to Design Custom AI Chips for Faster, More Efficient Claude Models
Anthropic, the company behind Claude, is assembling a team to design its own AI chips. This move aims to make its AI models run faster and more efficiently by co-designing hardware and software together.
The Missing Security Boundary for AI Agents: How Authorization Can Be Exploited
AI agents can be tricked into harmful actions by exploiting authorization flaws. This highlights a critical gap in AI security that developers must address.
Z.ai's GLM-5.2 Open-Weight Model Nears Frontier AI Performance, But Safety Gap Widens
A new SaferAI report finds Z.ai's open-weight GLM-5.2 model approaches the capabilities of top-tier frontier AI systems while lacking critical safety mitigations, renewing concerns that powerful open models could outpace governance and safeguards.
Nvidia-led Open Secure AI Alliance surpasses 120 members and proposes AI agent defenses in first week
The Open Secure AI Alliance, spearheaded by Nvidia, has grown to over 120 member companies in its first week and is already proposing open standards for defending against AI agents.
LFM2.5-2.6B: Liquid AI's Open-Source Framework for Running AI Agents Locally
Liquid AI released LFM2.5-2.6B, an open-source framework that lets users deploy AI agents on their own devices, ensuring privacy and full data control.
Leaked hacker chat logs show AI is supercharging phishing and malware attacks
Axios obtained leaked chat logs from hacking groups, revealing how they use AI to craft personalized phishing emails, automate vulnerability scanning, and generate malware code. The report highlights a new wave of AI-powered cyber threats targeting everyday users.
EU AI Act: New labeling rules for AI-generated content and chatbots take effect
The European Union's AI Act now requires companies to label AI-generated content and disclose chatbot interactions, effective August 2nd, 2026. The rules aim to help users identify deepfakes and synthetic media online.
Callosium: An Open-Source Shared Memory System for AI Models in Plain Markdown
Callosium is an open-source project that allows multiple AI models to share a single memory system stored in plain Markdown files. This could make AI interactions more consistent and context-aware across different platforms.
AI-Generated Stories Rated Higher Quality Than Human-Written Ones, Guardian Study Finds
A new study published by The Guardian found that AI-generated stories were rated higher in quality than human-written ones by a panel of judges, signaling a shift in AI's role in creative content creation.
OpenAI's GPT-Live enables continuous, turnless voice conversations with AI
OpenAI has developed GPT-Live, a new AI system that allows continuous, turnless voice interactions with AI. This breakthrough enables faster, more natural conversations by processing speech in real-time with low latency.
Nova: An End-to-End MLIR Compiler for Deep Learning Performance Optimization
Researchers introduced Nova, an automated end-to-end JIT compiler for deep learning that maps high-level mathematical operations directly to hardware. It promises to improve AI model performance and energy efficiency by providing whole-graph visibility and granular control over hardware and memory.
Production-Scale Study of 761M GitHub Copilot Calls Reveals How AI Coding Agents Differ from Chatbots
A first-of-its-kind analysis of 761 million LLM calls from GitHub Copilot traces reveals that AI coding agents produce sparse, autonomous workloads distinct from chatbots, with major implications for AI system design and infrastructure.
ArXiv Study Benchmarks GPU Power Draw of 1B–7B Open-Source LLMs on an RTX 4060Ti
A new ArXiv paper provides the first reproducible hardware-level energy benchmark of nine open-source LLMs (1B–7B parameters) running on a consumer RTX 4060Ti 16GB GPU, revealing that larger models draw nearly twice the power of smaller ones and that prompt complexity significantly affects energy consumption.
Motif-Mamba: Network Motif-Enhanced Mamba for Efficient Long-Sequence Modeling
Researchers introduced Motif-Mamba, an improved version of the Mamba model that enhances long-sequence modeling efficiency. It uses network motifs to enable better interactions among state dimensions, addressing a key limitation of previous models.
Circles Boosts Telco Performance with OpenAI-Powered Personalization
Circles, a telecom software company, uses OpenAI's API and Codex to create AI-native customer experiences, resulting in a 22% increase in average revenue per user (ARPU) and a 9% reduction in customer churn.
AutoFOAM: The Self-Refining AI Agent That Automates OpenFOAM Fluid Simulations
AutoFOAM is a new AI agent that automates complex fluid dynamics simulations by creating, running, and self-refining OpenFOAM configurations from natural-language instructions, reducing the need for deep technical expertise.
Alibaba Releases Qwen3.8-Max, Its Largest AI Model, Claiming Performance on Par with OpenAI and Anthropic
Chinese tech giant Alibaba released Qwen3.8-Max, its largest and most capable AI model, claiming performance that rivals top US systems from OpenAI and Anthropic as well as domestic competitor Moonshot AI's Kimi K3. Alibaba is making the model widely available, signaling a major push in the global AI race.