Generative AI news and analysis on Web Pulse: 12998 curated articles tracking the latest Generative AI developments, tools, and research, updated continuously from vetted sources.
Security researchers have identified a prompt injection vulnerability in Google Gemini's voice assistant that allows attackers to embed malicious instructions within ordinary smartphone notifications. The assistant proce…
Enterprise buying now involves multiple stakeholders who research independently before contacting vendors, with Gartner's 2025 research finding that 61% of B2B buyers prefer a rep-free experience. AI agents are being dep…
A new population-scale synthetic dataset for El Salvador has been released, containing over 8,000 multilingual, region-specific personas with detailed demographic, professional, and cultural attributes. The dataset, whic…
A developer rebuilt HerRights, a legal rights tool for vulnerable women, from a broken 24-hour hackathon prototype into a production-ready accessibility application using GitHub Copilot. The original app suffered from 10…
Apple silicon's integration of CPU, GPU, Neural Engine, and unified memory has turned Mac minis and Mac Studios into the dominant hardware platform for running personal AI agents, driving backorders of four to five month…
AI is rapidly reducing the cost of software development, shifting the role of developers from specialized engineers to generalists akin to home builders. As coding becomes less scarce, software companies now compete on d…
GitHub Copilot switched to metered token-based billing on June 1, 2026, prompting a wave of developers to threaten leaving the platform. The change, which could cost individual users hundreds of euros more per month, end…
Broadcom CEO Hock Tan announced during the company’s Q2 fiscal 2026 earnings call that order visibility now extends to 2028, driven by $30 billion in quarterly AI semiconductor bookings. The chipmaker posted record reven…
Brands are increasingly posting on Reddit to manipulate AI search results, with a subreddit dedicated to biohacking recently banning peptide-related content due to corporate astroturfing. Meanwhile, the rise of principal…
Researchers have developed a training-free diffusion model that generates new images matching the internal patch structure of a single reference image, eliminating the need for computationally expensive neural network tr…
Researchers have developed POLARIS-9B, a 9-billion-parameter language model that generates long-form stories with improved quality and length adherence, using a novel training recipe that combines a frontier LLM judge wi…
A mixed-methods study of mathematicians using AI for proof formalization found that users prefer AI assistance that preserves human control over proof discovery. In a controlled user study, participants achieved higher f…
Researchers have developed UniCanvas, a diffusion-based unified model that generates interleaved text and images within a single pixel canvas. The model embeds language as visual patterns during image synthesis, overcomi…
A new arXiv paper presents a data-centric taxonomy of 3D vision, connecting geometric representations, datasets, learning frameworks, and applications into a single conceptual map. The work analyzes principal 3D data str…
Researchers found that a random forest classifier using interpretable linguistic features—such as lexical diversity, readability, and emotional intensity—achieved near-perfect performance in detecting AI-generated fake n…
Researchers have developed a diffusion model that separates reflections from a single image by simultaneously generating transmission and reflection layers through a unified framework. The model introduces a cross-layer …
A large-scale empirical study analyzing 284 linguistic features across 27 large language models and ten text domains found that classifiers based solely on these features can reliably distinguish AI-generated from human-…
Researchers introduced EVID-Bench, a benchmark requiring systems to search the open web for related videos and identify false information through cross-video comparison, as video misinformation increasingly relies on evi…
Researchers have developed AXON, a training-free module that improves the quality-latency trade-off in discrete diffusion language models by selecting which masked tokens to reveal based on their ability to support later…
A developer has outlined a practical RAG evaluation checklist for AI SaaS products, emphasizing that retrieval-augmented generation systems can fail in subtle ways—such as sounding correct while teaching wrong workflows—…