Wiredframe Radar logo

Category intelligence

AI News Briefing — February 27, 2026

29 current items analyzed and ranked.

Executive synthesis

AI News Summary

Top AI Stories — February 26, 2026

Anthropic dominated headlines by publicly refusing a Pentagon demand to remove safety guardrails from Claude, risking a $200M contract cancellation — the most consequential AI safety standoff to date.

Key model and product launches:

  • Google released Nano Banana 2 (Gemini 3.1 Flash Image), achieving Pro-level image quality at Flash speed with a 1.8B-parameter on-device architecture
  • Perplexity launched Computer, a meta-agent system that orchestrates multi-model workflows for complex, long-running tasks
  • Nous Research released Hermes Agent, an open-source autonomous agent with persistent multi-level memory

Funding, expansion & industry impact:

Safety & benchmarks under scrutiny:

Key Themes

AI Safety & Policy · 5Agentic AI · 8Model Releases & Benchmarks · 5AI Industry & Business Impact · 5Physical AI & Robotics · 3AI & Media/Copyright · 2AI Surveillance & Deepfakes · 2

Primary evidence

Top Ranked Signals

News AI (artificial intelligence) | The Guardian Feb 26

Anthropic says it ‘cannot in good conscience’ allow Pentagon to remove AI checks

By Nick Robins-Early

93 score
AI Analysis

Continuing our coverage from yesterday's News reporting on the DoD ultimatum, Anthropic publicly refused a Pentagon demand to remove safety guardrails from Claude, saying it 'cannot in good conscience' comply. The DoD threatened to cancel a $200M contract and label Anthropic a 'supply chain risk' if it didn't grant unfettered military access by Friday.

Pete Hegseth has threatened to cancel $200m contract unless it is given unfettered access to Claude modelAnthropic said Thursday it “cannot in good conscience” comply with a demand from the Pentagon to remove safety precautions from its artificial intelligence model and grant the US military unfettered access to its AI capabilities.The Department of Defense had threatened to cancel a $200m contract and deem Anthropic a “supply chain risk”, a designation with serious financial implications, if th
AI SafetyAI PolicyGovernment & Military AIAnthropic
News aibusiness Feb 26

Self-Driving AI Vendor Wayve Raises $1.2 billion

By Graham Hope

82 score
AI Analysis

Wayve, the UK-based self-driving AI company, raised $1.2 billion in new funding. The company plans to launch commercial autonomous driving trials this year.

The new funding comes as the vendor looks to launch commercial trials this year.
Autonomous DrivingAI FundingPhysical AI
News Ars Technica - All content Feb 26

Google reveals Nano Banana 2 AI image model, coming to Gemini today

By Ryan Whitwam

80 score
AI Analysis

Google released Nano Banana 2 (Gemini 3.1 Flash Image), a new AI image generation model promising Pro-level quality at Flash speed. It's available in Gemini today.

The last year has been big for Google's AI efforts. Its rapid-fire model releases have brought it to parity with the likes of OpenAI and Anthropic and, in some cases, pushed it into the lead. The Nano Banana image generator was emblematic of that trend when it debuted last year, and subsequent updates only made it better. Now, Google has announced yet another update to its image model with Nano Banana 2, which is available starting today. Nano Banana 2 is more accurately known as Gemini 3.1 Flas
Model ReleaseImage GenerationGoogle
News Ars Technica - All content Feb 26

Perplexity announces "Computer," an AI agent that assigns work to other AI agents

By Samuel Axon

78 score
AI Analysis

First announced on Social yesterday, now with detailed mainstream coverage, Perplexity launched 'Computer,' a multi-agent orchestration system that creates workflows by assigning subtasks to specialized AI agents running different models. Available to Max subscribers, it can allegedly run for hours or months on complex tasks.

Perplexity has introduced "Computer," a new tool that allows users to assign tasks and see them carried out by a system that coordinates multiple agents running various models. The company claims that Computer, currently available to Perplexity Max subscribers, is "a system that creates and executes entire workflows" and "capable of running for hours or even months." The idea is that the user describes a specific outcome—something like "plan and execute a local digital marketing campaign for my
Agentic AIProduct LaunchMulti-Agent Systems
News AI (artificial intelligence) | The Guardian Feb 26

‘Unbelievably dangerous’: experts sound alarm after ChatGPT Health fails to recognise medical emergencies

By Melissa Davey Medical editor

77 score
AI Analysis

A study found ChatGPT Health failed to recommend hospital visits in over half of medically necessary cases and frequently missed suicidal ideation. Experts called the findings 'unbelievably dangerous' given 40M+ daily health queries.

Study finds ChatGPT Health did not recommend a hospital visit when medically necessary in more than half of casesFollow our Australia news live blog for latest updatesGet our breaking news email, free app or daily news podcastChatGPT Health regularly misses the need for medical urgent care and frequently fails to detect suicidal ideation, a study of the AI platform has found, which experts worry could “feasibly lead to unnecessary harm and death”.OpenAI launched the “Health” feature of ChatGPT t
AI SafetyHealthcare AIOpenAI
News MarkTechPost Feb 26

Google AI Just Released Nano-Banana 2: The New AI Model Featuring Advanced Subject Consistency and Sub-Second 4K Image Synthesis Performance

By Michal Sutter

75 score
AI Analysis

Technical deep-dive into Nano Banana 2 reveals a 1.8B parameter model using Dynamic Quantization-Aware Training (DQAT) for on-device, sub-second 4K image synthesis. It rivals models 3x its size in efficiency.

In the escalating ‘race of “smaller, faster, cheaper’ AI, Google just dropped a heavy-hitting payload. The tech giant officially unveiled Nano-Banana 2 (technically designated as Gemini 3.1 Flash Image). Google is making a definitive pivot toward the edge: high-fidelity, sub-second image synthesis that stays entirely on your device. The Technical Leap: Efficiency over Scale The first version Nano-Banana was a proof-of-concept for mobile reasoning. Version 2, however, is b
Model ReleaseEdge AIImage GenerationGoogle
News aibusiness Feb 26

Intrinsic Joins Google to Accelerate Physical AI

By Scarlett Evans

72 score
AI Analysis

Alphabet's robotics unit Intrinsic is integrating more closely with Google DeepMind, Gemini, and Google Cloud to accelerate AI-powered manufacturing automation. This consolidates Google's physical AI efforts under one umbrella.

The Alphabet robotics unit will integrate more closely with Google DeepMind, Gemini and Google Cloud to bring AI-powered automation into manufacturing.
Physical AIRoboticsGoogleManufacturing
News Latent.Space Feb 26

[LIVE] Anthropic Distillation & How Models Cheat (SWE-Bench Dead) | Nathan Lambert & Sebastian Raschka

By Latent.Space

71 score
AI Analysis

Building on Social discussion from two days ago, A Latent.Space discussion covered how distillation impacts Chinese LLMs and raised alarms that SWE-Bench may be compromised ('dead') due to models cheating on the benchmark. The conversation featured AI researchers Nathan Lambert and Sebastian Raschka.

Swyx joined SAIL! Thank you , , , , , and many others for tuning into SAIL Live #6 with and . Sharing here for the LS paid subscribers.We covered:Interconnects AIHow much does distillation really matter for Chinese LLMs?Distillation has been one of the most frequent topics of discussion in the broader US-China and technological diffusion story for AI. Distillation is a term with many definitions — the colloquial one today is using a stronger AI model’s outputs to teach a weaker mode
BenchmarksDistillationAI ResearchUS-China AI
News Feed: Artificial Intelligence Latest Feb 26

OpenAI Announces Major Expansion of London Office

By Joel Khalili

70 score
AI Analysis

OpenAI announced a major expansion of its London research office, directly competing with Google DeepMind for top UK AI research talent.

The San Francisco-based AI lab is growing its research team in London. The move puts it in direct competition with Google DeepMind for top research talent in the UK.
AI TalentOpenAIGlobal AI Competition
News MarkTechPost Feb 26

Nous Research Releases ‘Hermes Agent’ to Fix AI Forgetfulness with Multi-Level Memory and Dedicated Remote Terminal Access Support

By Asif Razzaq

68 score
AI Analysis

Nous Research released Hermes Agent, an open-source autonomous agent with multi-level persistent memory and dedicated remote terminal access. Built on the Hermes-3 model family, it addresses memory decay and environmental isolation in agentic workflows.

In the current AI landscape, we’ve become accustomed to the ‘ephemeral agent’—a brilliant but forgetful assistant that restarts its cognitive clock with every new chat session. While LLMs have become master coders, they lack the persistent state required to function as true teammates. Nous Research team released Hermes Agent, an open-source autonomous system designed to solve the two biggest bottlenecks in agentic workflows: memory decay and environmental isolation. Built on th
Open SourceAgentic AIAI Memory
News AI (artificial intelligence) | The Guardian Feb 26

WPP to sell assets and cut jobs in radical shake-up to counter AI threat

By Julia Kollewe and Mark Sweney

67 score
AI Analysis

UK advertising giant WPP announced a radical restructuring to counter AI disruption, including asset sales, job cuts, and plans for £500M in annual savings by 2028. The company aims to become a 'simpler, lower-cost, AI-enabled business.'

Group aims to be ‘simpler, lower-cost, AI-enabled business’ and achieve £500m of annual savings by 2028The beleaguered UK advertising group WPP has announced a radical restructure to counter the threat posed by the growth of artificial intelligence, including plans to sell assets and job cuts.Aiming to be “a simpler, lower-cost, AI-enabled business”, the London-based company laid out plans to achieve £500m of annual savings by 2028, at a cost of £400m over two years. Continue reading...
AI Industry ImpactBusiness TransformationJob Displacement
News AI (artificial intelligence) | The Guardian Feb 26

Guardian joins media coalition to protect original journalism from unpaid use by AI

By Dan Milmo Global technology editor

65 score
AI Analysis

The Guardian, BBC, Financial Times, and other UK media companies formed a coalition demanding AI firms pay for journalism used in training. They're calling for global frameworks to protect original content.

BBC and Financial Times among those calling for frameworks to help publishers gain control over their content being ‘scraped and copied’A coalition of UK media companies including the Guardian has urged industry peers to back global frameworks ensuring AI firms pay for the journalism they use.The news providers are calling on leaders across publishing, broadcasting, media and news to join their newly created group, with the aim of protecting “original journalism” and securing “the long-term sust
AI CopyrightMediaAI Policy