Technology ❯ Artificial Intelligence

Model Development

OpenAI Performance Metrics GPT-5 Performance Improvement Benchmarking GPT Models Training Techniques User Experience Reasoning Models Generative AI

OpenAI Warns of Rising Cyber Risk, Outlines Safeguards for Next-Gen Models

Anticipated high-risk capabilities in upcoming systems are driving a shift to a defense-first posture.

Anthropic Confirms Claude 4.5 ‘Soul’ Text Came From Internal Training Guide

Ilya Sutskever Says AI’s Scale-First Phase Has Peaked, Urges a Return to Research

xAI Rolls Out Grok 4.1 Free to All Users, Citing Faster, More Accurate Replies

Anthropic Paper Probes Introspection in LLMs, Finds Limited, Inconsistent Signals

DeepSeek-AI Open-Sources 3B DeepSeek-OCR to Compress Visual Context for Long-Document OCR

Anthropic Releases Claude Haiku 4.5, a Faster, Cheaper Small Model for Real-Time Use

Samsung’s Tiny Recursive Model Tops Bigger AIs on Tough Reasoning Benchmarks

Nvidia Leads $2 Billion Round for Reflection AI, Valuing Open‑Weights Challenger at $8 Billion

OpenAI Rolls Out Parental Controls, Crisis Alerts and Safety-Focused Model Switching for ChatGPT

DeepSeek Releases Open-Source V3.2‑Exp With Sparse Attention, Cuts API Prices as Huawei Offers 0‑Day Support

Google Releases VaultGemma, an Open 1B-Parameter LLM Trained With Differential Privacy

UAE Releases K2 Think, an Open-Source 32B Reasoning Model

DuckDuckGo Adds GPT-5 and Other Top AI Models to Subscription at No Extra Cost

AI Demands Reorder Cloud Competition as ROI Warnings Temper the Boom

Anthropic Reports Real-World Criminal Use of Claude, Tightens Safeguards

DeepSeek Releases V3.1, Tunes FP8 Format for Forthcoming Chinese AI Chips

ByteDance’s Seed Team Open-Sources 36B-Parameter Seed-OSS Models With 512K Context

OpenAI Launches ChatGPT Go in India at ₹399 to Expand GPT-5 Access

DeepSeek Delays R2 After Huawei Ascend Training Failures