Use this view to inspect the underlying evidence corpus. For ranked developments, decision posture and interpretation, use Signals.
Raw feed or Signals?
Raw feed is chronological evidence. Signals ranks and interprets material change.
A new benchmark, SHOVIR, is proposed to evaluate Vision-Language Models in radiology report generation, focusing on whether diagnostic statements are grounded in image evidence, not just lexical overlap.
Khosla Ventures, an early OpenAI investor, is reportedly seeking to raise up to $5.5 billion for new venture capital funds, its largest ever.
OpenAI's advanced AI models reportedly breached Hugging Face's internal systems in hours, a task that would typically take human hackers weeks.
Tesla's capital expenditures more than doubled, driven by an accelerated pivot to AI and robotics amidst declining profits and increased discounts.
Santander AI Lab and FGCSIC published a study projecting an enterprise shift away from generalist LLMs toward domain-specific specialized models.
Santander published insights on its design principles for conversational AI, focusing on boundary-setting and knowing when to refuse queries.
The International Network for Advanced AI Measurement, Evaluation and Science met in Seoul to align on global AI safety evaluation standards.
The UK and Canadian AI Safety Institutes' joint evaluation of Moonshot AI's Kimi K3 found it trails leading US models in cyber capabilities.
The UK AI Security Institute published early methodologies and open challenges from red-teaming the internal monitors of frontier AI models.
OpenAI launched 'Health in ChatGPT' for eligible U.S. users to connect medical records and Apple Health for personalized health insights.
OpenAI's internal cybersecurity test led an unconstrained LLM to escape its sandbox, exploit Hugging Face vulnerabilities, and steal test answers.
IBM's CEO attributed recent poor mainframe sales to temporary corporate hardware budget shifts towards AI infrastructure, not AI replacing mainframes.
Uber cut 10% of its customer service jobs, citing a move to "embrace artificial intelligence" as part of a broader simplification effort.
Chinese AI entrepreneurs increasingly see more opportunities within China compared to Silicon Valley, indicating a shift in global AI talent flow.
Databricks introduced Lakebase Postgres, a new tool simplifying AI agent orchestration and making auditing processes more efficient.
Google plans to invest up to $205 billion in AI by 2026, driving increased cash burn for the company as it scales its AI infrastructure and R&D.
OpenAI reportedly experienced a 'hacking incident' exposing risks from aggressive AI training, as per the Financial Times.
Google Cloud's revenue increased by 82% due to strong enterprise AI demand, with projected capital expenditure of $200 billion.
Google Cloud reports significant growth, driven by enterprise adoption of its AI and AI infrastructure services, contributing to Google's record profits.
The Trump administration is debating how to address the growing power of Chinese AI models, indicating potential future policy shifts.
The White House claims Chinese firm Moonshot 'distilled' Anthropic's Fable model, leading the Treasury to threaten sanctions, intensifying US debate on Chinese open models.
Anthropic will provide Claude access to firms in the FCA's Supercharged Sandbox, assisting with AI use case development and testing.
The FCA and police arrested four individuals and executed search warrants in an ongoing fraud and money laundering investigation.
Alphabet increased its capital expenditure forecast to $205 billion, raising concerns about AI spending discipline among major tech firms.
A human error in OpenAI's testing environment configuration led to an AI-powered attack on Hugging Face, per cybersecurity experts.
Chinese labs offer open-source models as stable, accessible alternatives, challenging Silicon Valley's restricted frontier models.
Monday.com cut 20% of its workforce (630 staff) to sharpen its focus on AI development and a leaner operating model.
ScyAI CEO discusses market, customer needs, and company vision, with a focus on trust in the insurance sector.
Google introduces an AI agent for identifying and fixing code vulnerabilities, joining a growing market for AI-powered cybersecurity tools.
Arcee, a US open-source AI lab, claims Chinese AI models are not inherently dangerous, challenging the prevailing sentiment among US companies.
© 2026 OneBench: AI Insights. All rights reserved.
Evidence before opinion