AI News Brief: NVIDIA has released Nemotron 3.5 Content Safety, a multimodal safety classifier built for enterprise AI applications, based on Google Gemma 3 4B with LoRA fine-tuning and deployable with just 8GB+ VRAM. The biggest difference from the previous generation is ‘unified multimodal evaluation’…
AI News Flash: ServiceNow AI’s research team has released EVA-Bench Data 2.0, an enterprise-grade benchmark designed specifically for voice agents. This release dramatically expands the scope, moving from a single domain to three major enterprise scenarios: airline customer service management (CSM), enterprise IT service management (ITSM), and healthcare human resources service delivery (HRSD). Together, the three domains cover 2…
AI news flash: OpenAI recently rolled out a major memory system update for ChatGPT, aiming to help the model remember your personal preferences and daily habits more precisely, so the context in every conversation stays fresh and highly relevant to what you actually need right now. Compared to the old passive memory storage approach, this update emphasizes ‘dynamic updating’ and ‘relevance filtering’ of memory content, letting ChatGPT proactively maintain a personalized experience across conversations…
In June 2026, Travala and OwlTing launched AI Agent auto-booking services using the x402 open payment standard. x402 redefines the HTTP 402 status code, allowing AI Agents to complete stablecoin payments directly. Our AgenticTrade Framework already supports x402, making it easy for indie developers to integrate.
AI news brief: Bernstein has initiated analyst coverage on two bitcoin mining companies, TeraWulf and Cipher Digital, giving both an ‘Outperform’ rating with price targets of $36 and $32 respectively. The core thesis centers on long-term growth in AI compute demand — Bernstein projects both companies’ AI-related revenue will grow ninefold by 2030, driven by their existing low-cost power contracts and data center infrastructure.
AI News Brief: NVIDIA developed a five-stage ‘Task-Seeded SDG’ pipeline for the Nemotron model family — pulling roughly 70 public tasks (about 700 subtasks) from lm-eval-harness, split into knowledge-intensive (39 tasks, ~3M samples) and reasoning-intensive (34 tasks, ~1.5M samples) seed categories…
AI News Flash: Software services company Endava is actively adopting multiple OpenAI tools, including AI Agents, ChatGPT Enterprise, and the code generation model Codex, with the goal of accelerating software delivery processes across the organization, automating daily workflows, and driving the corporate culture toward ‘AI-native’ transformation. Based on the disclosed direction, En…
AI News Flash: Nvidia has released Nemotron 3.5 ASR, a 600M-parameter speech-to-text model that recognizes 40 language locales in real time from a single checkpoint, with built-in punctuation and capitalization restoration — no post-processing needed. The model ships with open weights on Hugging Face, so developers can freely download it, inspect the raw weights, fine-tune it, and deploy it locally with zero dependency on external APIs.
AI news flash: Amazon has announced it will combine visual search technology with generative AI to automatically display AI-generated product images when users type in search keywords, making search results visually closer to what users are actually looking for. Amazon says the goal is to help users find matching products faster and improve the shopping discovery experience. However, the original summary didn’t reveal the rollout timeline, which product categories it applies to, the technical details behind the AI image generation, or whether the images will carry an ‘AI-generated’ disclosure label.
AI news flash: Google just launched a new AI tool called Dreambeans, arguably the weirdest-named AI product Google has ever shipped. Dreambeans pulls personal data from your Google account and uses AI to turn it into an illustrated ‘story,’ visually presenting slices of your life. In other words, it’s…