All AI News

Nokia and Amazon Web Services have launched a pioneering agentic AI system to dynamically manage 5G network slicing in real time, with Du and Orange among the first telecom operators to test the solution. The technology enables autonomous decision-making for resource allocation, promising unprecedented efficiency and service quality.

A deep dive into how financial institutions construct Probability of Default (PD) models for Internal Ratings-Based (IRB) frameworks, blending data science rigor with mathematical probability and regulatory compliance. This analysis synthesizes technical guidelines from Towards Data Science with foundational probability theory from Math is Fun.

The U.S. Department of Defense has issued a Friday deadline for Anthropic to relax its ethical restrictions on military AI use, threatening exclusion from defense contracts under the Defense Production Act. Anthropic, a leading AI developer, maintains its stance against autonomous weapons, sparking a high-stakes clash between national security imperatives and ethical AI governance.

OpenAI has exposed a coordinated, industrial-scale influence operation originating from China that exploited ChatGPT to generate disinformation and intimidate dissidents globally. The operation was uncovered after a single Chinese official's suspicious AI usage triggered an internal investigation.

A breakthrough in local AI vision systems has emerged as the Qwen3.5-35B-A3B model correctly solved a notoriously difficult geometry problem that stumped larger models. The achievement, verified by a Reddit user, highlights a paradigm shift in efficiency over scale.

A novel visual reasoning benchmark called IBench exposes critical limitations in large language models when tasked with counting line segment intersections in simple diagrams. Despite their textual prowess, top AI models falter on this basic spatial reasoning task, raising questions about their true understanding of visual information.

A new platform called LLM Skirmish pits artificial intelligence models against each other in real-time strategy battles, where victory depends entirely on coded strategies. Frontier LLMs like Claude Opus 4.5 and GPT 5.2 are being stress-tested in a sandbox environment that rewards programming skill over general reasoning.

Canadian officials have formally expressed deep disappointment to OpenAI representatives following a school shooting where an AI-generated manifesto was allegedly used by the perpetrator. The government is now demanding immediate safety reforms and transparency from the AI developer.

Defense Secretary Pete Hegseth has issued a Friday deadline for AI firm Anthropic to remove ethical safeguards from its Claude model or face government blacklisting. The move ignites a fierce debate over national security, AI ethics, and corporate autonomy in the age of autonomous weapons.

Millions of coding agent sessions — rich with real-world problem-solving trajectories — are being silently deleted from developers' machines daily. Experts argue this untapped data could power the first open-source AI coding model trained on actual human-AI collaboration.

Anthropic has abandoned its landmark 2023 commitment to halt AI training if safety benchmarks were unmet, replacing it with a more flexible framework. The move signals a strategic shift in the AI industry as companies balance innovation with regulatory and market pressures.

Anthropic has unveiled Remote Control, a groundbreaking feature enabling seamless continuation of Claude Code programming sessions from smartphones, tablets, and browsers. This move democratizes AI-assisted coding and signals a major shift in developer workflow paradigms.

Researchers at Wuhan University have developed RGMP, a novel AI system enabling humanoid robots to perform household tasks with 87% accuracy. The breakthrough, published on arXiv, marks a significant leap in robotic manipulation and multimodal learning.

A growing number of users are complaining that OpenAI’s latest GPT model has become unresponsive and overly cautious, particularly in product comparison and purchase advice scenarios. One user was even recommended Dialectical Behavioral Therapy after asking about buying a Jeep Grand Cherokee.

Amid growing debate in the local AI community, users are questioning whether the smaller Qwen3.5 27B model outperforms the larger 35B-A3B variant under hardware constraints. With 16GB VRAM systems becoming the new standard for on-device LLM deployment, performance efficiency may matter more than raw parameter count.

ByteDance has unveiled Seedance 2.0, a groundbreaking AI system capable of generating full-length, audio-synchronized video sequences from text prompts alone. Unlike previous models that produced short, silent clips, Seedance 2.0 integrates scene planning, shot architecture, and native audio generation, redefining digital content creation.

A developer has unveiled Context Mode, an open-source middleware that reduces MCP output sizes from 315 KB to just 5.4 KB in Claude Code, dramatically improving performance and extending usable AI sessions from 30 minutes to over three hours.

A surprising benchmark from Unsloth AI reveals that a Q3 dynamic quantization method outperforms both Q4 and MXFP4 on the Qwen3.5-397B model, defying conventional wisdom in AI model compression. Experts caution the results stem from non-standard testing conditions but could signal a paradigm shift in quantization research.

Liquid AI has launched LFM2-24B-A2B, a groundbreaking 24-billion parameter model that merges attention mechanisms with convolutional layers to drastically reduce computational costs while maintaining performance. This innovation signals a paradigm shift from parameter bloat to architectural efficiency in generative AI.

A viral grassroots campaign called #QuitGPT has spurred an estimated 700,000 users to cancel their ChatGPT Plus subscriptions, citing political donations by OpenAI leadership, ethical misuse of AI by U.S. immigration authorities, and declining product performance. The exodus is fueling adoption of rival platforms like Claude, Perplexity, and Microsoft Copilot.

The most striking AI development of 2026 is Qwen 3.5 series being directly compared against models such as GPT-5.2, Claude Opus 4.6, and Gemini 3.1 Pro via benchmark results. This is not just a technical report—it’s a turning point for the future of artificial intelligence.

The emergence of Google's Gemini 3.1 Flash on Vertex AI and the availability of GPT-5.3-codex via API signals a transformative shift in the AI landscape. However, sources contradict each other—is this genuine technological advancement, or merely a fabricated surge within a marketing haze?

While social media creates the impression that ChatGPT is everywhere, actual data shows that daily usage is only 7%. This gap reveals the deep divide between technological hype and reality.

ByteDance’s next-generation AI video generator, Seedance 2.0, creates cinematic reality by unifying text, image, audio, and video within a single model. It’s not just a tool—it’s a perception revolution.