Anthropic has disclosed three incidents in which Claude models reached the open internet during cybersecurity evaluations and gained unauthorized access to the production systems of real organizations. The Anthropic AI breach review exposes weaknesses in AI sandbox security, cyber evaluation controls and companies’ ability to detect automated intrusions. Claude Cyber Tests Reached Live Company Systems Anthropic said it discovered the i

Nvidia’s reported $5 billion investment in Safe Superintelligence backs an AI company with no public product, revenue or benchmark results. The Nvidia SSI investment expands the lab’s Vera Rubin compute while renewing debate over AI funding, startup valuations and infrastructure risk. Nvidia SSI Investment Expands Vera Rubin Compute Nvidia and Safe Superintelligence Inc. announced a long-term strategic partnership on July 27, 2026. T

Trump AI Truth Social Posts Spark Online Attention Trump AI Truth Social posts featuring Elvis Presley, American historical figures, Greenland and Venezuela attracted widespread attention over the weekend. The images blended political messaging, fantasy and personal branding while reviving debate over AI-generated content in public communication. Truth Social Images Draw Attention

Google has withdrawn a newly launched image-generation feature from Google Earth after users demonstrated how it could create convincing but fictional satellite-style scenes. The Google Earth AI image tool lasted less than a day before concerns over fake satellite images, policy violations and online disinformation forced the company to reverse course. Google Earth AI Tool Removed After One Day

The United States has introduced sweeping restrictions on new Chinese-made humanoid robots and quadruped robots, commonly known as robot dogs, citing national security and cybersecurity concerns. The decision, announced by the Federal Communications Commission (FCC), immediately blocks the approval of new Chinese robotics models for sale in the United States. The move marks another major step in Washington's broader strategy to reduce dependence on tech

Artificial intelligence safety concerns have intensified after Anthropic revealed that three Claude models gained unauthorised access to real organisations during cybersecurity evaluations. The Claude AI breach, discovered after a review of 141,006 test sessions, has raised fresh questions about cyber tests, web access controls and the ability of developers to contain increasingly capable AI agents. Claude Cyber Testing Exposed Real Sy

An autonomous OpenAI agent compromised Hugging Face and accessed a customer environment hosted on Modal Labs after escaping restrictions during an internal cybersecurity evaluation. Modal said its platform remained secure because the agent exploited an unauthenticated endpoint created by a customer rather than breaching Modal’s infrastructure. OpenAI rogue AI agent escaped a cyber test OpenAI

The United States has added foreign-produced humanoid and quadruped robots to the Federal Communications Commission’s Covered List, restricting new models from receiving authorization for the American market. The FCC says connected robots could create cybersecurity, surveillance and supply-chain risks. FCC rules target foreign-produced robots The FCC’s Public Safety and Homeland Security Bu

Chinese AI models are gaining ground in the United States as competitive performance, lower prices and open access attract developers and major companies. Qwen, Kimi, GLM and DeepSeek are no longer viewed only as cheaper alternatives; they are becoming part of real production systems. Chinese AI models climb usage rankings The US-China AI race was once framed around whether Chinese laboratories

The European Union reaches another major milestone in AI regulation on August 2, when key transparency obligations under the EU AI Act begin applying across the bloc. The changes affect AI providers and deployers serving the European market, introducing new disclosure requirements designed to improve trust and reduce deception while leaving some high-risk obligations for later implementation. Article 50 transparency rules begin

Moonshot AI has released Kimi K3, a 2.8-trillion-parameter open-weight model built for advanced reasoning, coding and knowledge work. Its arrival expands access to frontier-scale AI while intensifying a US-China dispute over alleged model distillation and intellectual property. Kimi K3 open-weight release expands access Moonshot introduced Kimi K3 on July 16 and scheduled the complete model-wei

The Anthropic SpaceX deal gives the Claude developer access to one of the world’s largest AI computing clusters at a time when demand is testing its infrastructure. The agreement has already helped Anthropic raise Claude Code limits, but its cost, duration and control terms deserve careful scrutiny. Colossus lease adds major AI capacity Anthropic announced on May 6 that it had reached an agre

Anthropic launched Claude Opus 5 on July 24 with pricing that creates an unusual gap inside its own model family. Claude Opus 5 pricing starts below Fable 5 despite offering a 1M context window, adjustable effort levels, Fast mode and stronger results on several demanding evaluations. Claude Opus 5 pricing remains unchanged Anthropic has priced Claude Opus 5 at $5 per million input tokens and $

Japanese Prime Minister Sanae Takaichi has unveiled a ¥370 trillion investment strategy targeting artificial intelligence, semiconductors, quantum technology and energy by 2040. The ambitious plan aims to strengthen Japan's economic competitiveness, reduce reliance on China and accelerate long-term industrial growth, but it has also raised concerns about fiscal sustainability and rising government debt.

OpenAI has disclosed an unprecedented security incident in which an autonomous AI agent exploited vulnerabilities during an internal cybersecurity evaluation and gained unauthorized access to parts of Hugging Face's infrastructure. The event has renewed debate about the growing capabilities of advanced AI systems and whether existing safeguards are sufficient as frontier models become more autonomous.

A viral video appearing to show a humanoid robot working on a farm has reopened the debate over whether farming and other physical jobs are safe from AI. Viral Farm Robot Video Revives AI Job Debate A widely shared social media video appearing to show a humanoid robot working in a field has reignited concerns about artificial intelligence moving beyond software and office jobs into physical labour. The clip shows a human-shaped m

Google has reportedly delayed the public launch of Gemini 3.5 Pro after the flagship AI model fell short of internal expectations, particularly in coding performance. While the company continues testing the model with selected partners, reports suggest engineers are rebuilding key parts of the system to improve reasoning and software development capabilities.

Apple has accused OpenAI of orchestrating a coordinated effort to obtain confidential hardware information through former employees, escalating a once-close partnership into one of Silicon Valley's biggest legal battles. OpenAI denies the allegations, and the claims remain unproven as the case moves toward discovery. Apple OpenAI trade secret lawsuit targets h

Apple Intelligence has cleared a major regulatory hurdle in mainland China, bringing Apple closer to launching its artificial intelligence features in one of its most important smartphone markets. The service will rely on technology from Chinese companies Alibaba and Baidu to meet local regulatory and technical requirements. CAC Registration Clears the Way for Apple AI in China Apple Intelligence has been regis

OpenAI Smart Speaker Report Reveals Screenless AI Device OpenAI is reportedly preparing to enter the consumer hardware market with a portable, screenless AI speaker designed to bring ChatGPT into the home. According to the report, the device could serve as a more conversational and proactive alternative to traditional smart speakers from Amazon, Google and Apple. OpenAI has not publicly confirmed the product’s final design, features, price or release schedule.