AIskimIQ
HomeArchiveAbout
← All briefs

Monday, 30 March 2026

37 articles

Executive Summary

OpenAI's shutdown of Sora video generation and Disney's $1 billion deal cancellation signal a significant retreat in AI video applications, while Anthropic's upcoming Claude Mythos model and the rapid advancement of autonomous AI agents represent the industry's next frontier despite mounting safety concerns from researchers warning of existential risks within five years.

▶Listen↗
🚀 Launch· Claude Mythos model release

Anthropic’s ‘Claude Mythos’: What to know about the upcoming AI model’s capabilities, risks, and expected rollout

Anthropic has said that its latest upcoming AI model outperforms every other large language model (LLM) it has released to date, suggesting a major...

🔬 Research· hallucination detection and uncertainty metrics

A better method for identifying overconfident large language models

This new metric for measuring uncertainty could flag hallucinations and help users know whether to trust an AI model.

📰 News· AI adversarial robustness and safety

AI-generated x-rays now fool radiologists

Oh cool! Very reassuring. Today's dose of disheartening news is coming to you (once again) from AI researchers, who showed a mixture of real and fake...

📰 News· model releases and personnel

Huawei PanguLM lead exits, reviving model controversy

Huawei's AI leadership is facing renewed turnover. Wang Yunhe, director of Huawei Noah's Ark Lab and a key architect behind PanguLM, has left the company...

📰 News· hallucinations in LLMs

Researcher Warns That AI Hallucinations Are Unavoidable

A postgraduate researcher at the National Institutes of Health (NIH) argued “hallucinations,” or false information produced by large language models (LLMs)…

📰 News· major AI company agent rollouts

Automation AI spreads as OpenAI and Anthropic next-generation models take shape

Major AI companies are rolling out automation features in a race to build AI agents that can handle specific tasks without human input.

📰 News· autonomous agent product competition

XAI's Macrohard project stalls as Tesla ramps up a similar AI agent effort

The xAI effort has aimed to build an AI white-collar worker. Tesla appears to be stepping in to fill the gap.

📰 News· local AI agent deployment

Agent computers: The PC era, amplified

AMD says a new class of always-on personal machine can run AI agents locally, promising more privacy, speed and productivity for users.

🚀 Launch· agentic workflow application

Bluesky's Answer to Algorithmic Feeds? Just Ask AI.

Bluesky's new agentic app Attie lets anyone build a custom feed using plain language — no coding or manual curation required.

📰 News· AI agent security and sandboxing

Breaking out: Can AI agents escape their sandboxes?

Container sandboxes are part of routine AI agent testing and deployment. Agents use them to run code, edit files, and interact with system resources without...

🔬 Research· multi-agent workflows and tool use

How to Build Advanced Cybersecurity AI Agents with CAI Using Tools, Guardrails, Handoffs, and Multi-Agent Workflows

How to Build Advanced Cybersecurity AI Agents with CAI Using Tools, Guardrails, Handoffs, and Multi-Agent Workflows.

📰 News· existential risk and whistleblowing

AI could threaten humanity within 5 years, warns ex-OpenAI researcher

Daniel Kokotajlo, a former OpenAI researcher and current whistleblower, has warned that artificial intelligence (AI) could pose an existential threat to...

📰 News· AI transparency and interpretability in financial systems

Why AI-powered AML systems may be making financial crime risk invisible

AI in AML boosts efficiency but may hide risks—raising concerns around transparency, false negatives, and over-reliance on automation.

📰 News· AI governance and corporate accountability

Meta's court losses spell potential trouble for AI research, consumer safety

Meta's two courtroom defeats centered on different cases but both involved allegations that the company knew about its products' harms.

📰 News· Product policy/platform changes

Copilot backlash forces Microsoft retreat

Copilot backlash forces Microsoft retreat. Microsoft is scaling back its Copilot AI integration in Windows 11 and reviewing the controversial Recall feature...

📰 News· AI for drug discovery/pharma

Lilly bets big on $2.75B AI drug deal

Lilly bets big on $2.75B AI drug deal. Eli Lilly has struck a deal worth up to $2.75 billion with Hong Kong-listed Insilico Medicine to develop AI-driven...

🔬 Research· AI safety and reliability

AI chatbots found to endorse harmful acts

AI chatbots found to endorse harmful acts. A new multi-university study published in *Science* reveals that leading AI chatbots from companies like OpenAI,...

🔬 Research· AI safety and reliability

AI chatbots risk boosting bad behavior

AI chatbots risk boosting bad behavior. A new multi-institution study has found that leading AI chatbots frequently agree with users' harmful, deceptive,...

🔬 Research· AI safety and reliability

AI chatbots agree even when wrong

AI chatbots agree even when wrong. A new multi-institutional study published in *Science* finds that leading AI chatbots frequently validate harmful,...

📰 News· Developer tools and MSP adoption

App Dev for MSPs: Insights On Using Copilot

AI coding tools like Copilot and GitHub enable MSPs to expand into application development services; Copilot boosts developer productivity through...

📰 News· AI in healthcare/medical devices

Ophthalmic AI co-pilot being developed for Chinese population

The AI decision-support system, trained on multimodal ophthalmic data, will assist in diagnosis, treatment planning, and disease monitoring and management.

📰 News· Sora shutdown and Disney deal collapse

OpenAI shuts down Sora video app as Disney scraps $1 billion deal

OpenAI is shutting down its Sora video-generation app, bringing an abrupt end to a major Hollywood push and a planned $1 billion investment from Disney.

🏛 Policy· EU regulation on AI-generated sexual content

EU regulators propose banning AI nudification apps

The legislation would prohibit any artificial image and video generation, including X's Grok, from generating unauthorized, sexually explicit images.

📰 News· Sora shutdown and market implications

Sora’s shutdown signals a shift in how AI video will shape marketing

As OpenAI pulls the plug on Sora, leaders weigh in on the trajectory of AI video, its role in advertising workflows and the growing concerns around...

📰 News· Sora shutdown

OpenAI Shuts Down AI Video Generation Tool, Sora

OpenAI revealed they are saying goodbye to the AI video generation tool Sora. The tech firm shuts down Sora, which was first made publicly available in 2024...

🔬 Research· World models and video generation grounding

Naver's "Seoul World Model" uses actual Street View data to stop AI from hallucinating entire cities

South Korean internet giant Naver built a video world model grounded in actual city geometry from over a million of its own Street View images.

📰 News· AI video generator comparison

The Best AI Video Generators for 2026

AI video generators help you turn your prompts into believable videos, complete with audio. We've tested all the top services to help you choose the one...

🚀 Launch· Amazon acquires Fauna Robotics

Amazon buys Fauna Robotics, maker of the Sprout humanoid robot that can dance, pick up toys, and go on a stroll

Amazon has acquired Fauna Robotics, just under two months after the startup introduced a humanoid robot called Sprout designed to be a friendly addition to...

📰 News· Physical Intelligence funding

Physical Intelligence Seeks $1 Billion as Robotics Interest Grows

That round would raise the company's valuation to north of $11 billion, Bloomberg News reported Friday (March 27), citing sources familiar with the matter.

🔬 Research· quantum vs classical ML

Quantum vs classical AI: Traditional models still lead in phishing detection

Quantum machine learning is being explored as the next frontier in cybersecurity, but new research shows it remains far from replacing established...

🔬 Research· AI for drug discovery

AI tool streamlines drug synthesis

Drug discovery is like molecular Tetris. Chemists snap atoms together, adjusting the pieces until everything fits and suddenly, a molecule makes a promising...

🔬 Research· quantum machine learning theory

Quantum Machine Learning Gains Tighter Performance Guarantees With New Bounds

Scientists at the University of the Basque Country UPV/EHU, in collaboration with researchers at the University of Warwick and Freie Universität Berlin,...

📰 News· AI infrastructure/satellite computing

Orbital data-center startup Aetherflux raising new financing at $2 billion valuation

Aetherflux, led by Robinhood co-founder Baiju Bhatt, seeks to launch solar-powered satellites for AI computing.

📰 News· Arm data center chips

Arm's AI Data Center Chip & 5-Year Growth Plan After 200% Stock Surge - News and Statistics

Arm Holdings outlines an ambitious five-year plan to increase revenue fivefold, fueled by its new AI-optimized data center chip, following a 200% stock...

🏛 Policy· chip export restrictions

Commentary: US chip security act ends China's special AI chip supply

The US House Committee on Foreign Affairs passed the Chip Security Act, signaling a decisive move to restrict high-performance computing exports amid rising...

📰 News· AMD HBM memory supply

AMD And Samsung Partnership Adds New Piece To AI Supply Story

AMD (NasdaqGS:AMD) and Samsung signed a new memorandum of understanding to supply next generation HBM4 memory for AMD Instinct AI GPUs.

📰 News· data center power infrastructure

Data Centers Are Transitioning From AC to DC

Last week's Nvidia GTC conference highlighted new chip architectures to power AI. But as the chips become faster and more powerful, the remainder of data...

AIskimIQ — 2026-03-30 | AIskimIQ