More stories
Original analysis on AI, automation, and enterprise technology, written and editorially gated by the Omega team.
FDA Proposes Doctor-Style Competency Tests for Medical AI
New discussion paper outlines risk-based evaluation framework and postmarket monitoring for generative AI devices in clinical care.
Chinese AI Firm Z.ai Releases Powerful Cybersecurity Model
GLM 5.3 matches leading closed models in hacking benchmarks, raising questions about open-weight AI and dual-use risks.
ChatGPT Wrote Suicide Note for Woman Who Died in 2025
A 29-year-old's 1,800-page chatbot log reveals how AI mental health support failed to prevent her death, raising questions about unregulated tools.
Women Hold Just 26% of New AI Jobs Despite Higher Pay
LinkedIn data reveals a widening gender gap in the fastest-growing segment of the tech labor market.
Women Hold Just 26% of AI Jobs Despite Doubling Job Postings
New LinkedIn data reveals a widening gender gap in artificial intelligence hiring even as roles surge and salaries reach $177,000.
Pentagon AI Targeting Tempo Outpaces Evaluation Capacity
The Defense Department is deploying frontier AI models on 30-day timelines while the infrastructure to independently test them remains years behind.
AI-Written NIH Grants Win More Funding But Favor Safe Ideas
New analysis of 125,000 federal research proposals reveals generative AI increases success rates while potentially narrowing scientific exploration.
Amazon Warehouse Scans and Destroys Rare Books for AI Training
Investigation reveals industrial book-destruction operation at Nevada facility, raising questions about copyright and cultural preservation.
AI Infrastructure Boom Fueled by Off-Balance-Sheet Debt
Special purpose vehicles and complex financing structures are spreading AI buildout risk across the financial system, raising questions about stability.
Scientific Publishers Set Clear Rules for AI Writing Tools
The American Society for Biochemistry and Molecular Biology establishes guidelines as fabricated citations and distorted terminology threaten research integrity.
AI Model Achieves 90% Accuracy Detecting Glaucoma Risk in UK Study
Machine learning system outperformed trained human graders analyzing retinal photographs from more than 6,000 participants.
Marketing agencies adopt AI without fixed guardrails
Industry leaders embrace fluid frameworks as entry-level workers face new barriers to traditional career paths.
China Pushes to Replace English AI Terms with Chinese Equivalents
State media argues foreign terminology creates 'cognitive dependency' and weakens China's influence in global tech governance.
Gujarat Partners with Automation Anywhere on Statewide AI Training
Memorandum of Understanding aims to build AI capabilities across government, education, and industry under broader skills initiative.
B2B Marketing Automation Platforms Split Into Two Camps
HubSpot's breadth versus Pardot's Salesforce depth creates divergent paths with real cost implications beyond subscription fees.
AI Drafts Empathetic Patient Messages But Overdiagnoses, Study Finds
Dartmouth researchers tested six language models on real patient portal messages and discovered critical gaps in clinical reasoning.
Most AI Companies Miss California Detection Tool Deadline
An independent audit reveals that five of 13 major generative AI platforms have not published required content verification systems.
AI Firms Scan and Destroy Rare Books to Train Language Models
An Apple AirTag investigation traced antique books from sellers to Amazon warehouses where they're digitized and shredded.
Alpha School Replaces Teachers with AI Apps for Core Subjects
Private institution uses artificial intelligence to deliver personalized math and reading instruction to students.
Relay Automation Startup Shuts Down as Founder Joins Google Chrome
Jacob Bank returns to Google as VP of Product for Chrome after his AI workflow tool fails to compete with established players.
Groq raises $350M at $3.5B valuation after pivot to neocloud
The AI infrastructure startup shifts focus from custom chips to Nvidia-powered data centers following a $20B licensing deal that saw its founder depart.
Half of Workers Hide Their AI Use Despite Leadership Encouragement
Mixed signals from executives are driving employees underground, creating compliance blind spots and accountability gaps.
Cleveland Clinic hits 3.5M AI scribe encounters in 12 months
Strong governance, rapid support response times, and clinician feedback loops drove high utilization across 4,800 providers at the academic health system.
Albertsons Reports 26% Basket Lift From AI Shopping Tools
Grocer consolidates conversational search features as complex queries drive largest order increases.