AI News Archive: May 9, 2026 — Part 2
Sourced from 500+ daily AI sources, scored by relevance.
- Tech is turning increasingly to religion in a quest to create ethical AI
Tech is turning increasingly to religion in a quest to create ethical AI Inquirer.com
- India's market underperformance driven by 'AI exposure gap', not domestic weakness: Motilal Oswal report
Markets such as South Korea, Taiwan and the United States have significantly outperformed due to their strong exposure to the global AI trade, report says
- Nvidia and IREN target 5 gigawatts of power. Here is how much the deal is worth.
The Santa Clara chipmaker is investing $2.1 billion into the partnership to secure power for large-scale enterprise and hyperscale AI factories.
- AI cuts wildlife tracking time from months to days
Artificial intelligence can dramatically speed up the painstaking work of tracking wildlife with remote cameras, cutting analysis time from months or even a year to just days while producing nearly the same scientific conclusions as humans.
- India Seeks to Accelerate Responsible AI Adoption in Healthcare
At a time when the AI startups are facing major lawsuits from users on a variety of topics ranging from hallucinating chatbots causing suicidal actions among teenagers to fraudulent counselling delivered to mental health patients, India is moving ahead with measures to ensure the adoption of responsible AI in the country’s healthcare systems. Towards this […] The post India Seeks to Accelerate Responsible AI Adoption in Healthcare appeared first on CXOToday.com .
- 'LLMs can infer private attributes from ad exposure alone': AI can tell a lot about you just based on the overall patterns of ads you see, without the need to access your browsing history or personal data — and not even a VPN can protect you
AI can infer personal traits from ad exposure patterns alone, turning everyday advertising streams into detailed profiles without direct access to private data
- Mozilla Used Claude Mythos to Find 271 Firefox Bugs — Almost No False Positives
Mozilla uses Claude Mythos to find 271 Firefox bugs with high accuracy
Score: 50🌐 MovesMay 9, 2026https://opentools.ai/news/mozilla-claude-mythos-271-firefox-vulnerabilities - This 22-Year-Old SJTU PhD Is Building Flapping Wing Robots — and Sequoia Just Backed Him
Yingkong Zhivi, founded by four Shanghai Jiao Tong University PhD students, has raised tens of millions of RMB in a Pre-A round led by Yuanhe Origin, becoming the world's first company focused on embodied intelligence flapping wing robots.
- Musk, Altman management styles under fire at OpenAI trial
Testimony in the OpenAI feud revealed criticisms of Elon Musk's technical competency and temper, while Sam Altman faced accusations of dishonesty and integrity deficiencies from former board members. Jurors are weighing the credibility of both tech leaders as the trial delves into OpenAI's history and leadership disputes.
- What I saw at the Musk-OpenAI trial: petty billionaires, protests and a stern judge
Showdown between Musk and Altman has rendered the world’s most wealthy comical under egalitarian eye of court For the past couple of weeks, on the fourth floor of a courthouse on a quiet street in downtown Oakland, the world’s richest man and one of the world’s most valuable startups have been at war over the future of artificial intelligence. Being one of the reporters in the room has felt like watching an updated, opposite-coast version of Tom Wolfe’s The Bonfire of the Vanities – ambition, ego, greed and the spectrum of social class on full display. The supporting cast has included Elon Musk fanboys, a stern judge and a who’s-who of Silicon Valley’s most influential people. Continue reading...
Score: 50🌐 MovesMay 9, 2026https://www.theguardian.com/technology/2026/may/09/elon-musk-sam-altman-openai-trial - The Secret Diary That Has Spilled Into the Musk vs. OpenAI Feud
The personal journal of OpenAI president Greg Brockman is now a character in the company’s battle with the world’s richest man—and the most human part of a trial between tech billionaires
Score: 50🌐 MovesMay 9, 2026https://www.wsj.com/tech/musk-openai-trial-greg-brockman-diary-journal-6950270e?mod=rss_Technology - Amazon Admits Its Flagship AI Coding Tool Isn’t Good Enough for Its Own Workers to Use
It's officially opening the floodgates. The post Amazon Admits Its Flagship AI Coding Tool Isn’t Good Enough for Its Own Workers to Use appeared first on Futurism .
Score: 49🌐 MovesMay 9, 2026https://futurism.com/artificial-intelligence/amazon-admits-ai-coding-tool-isnt-good-enough - TikTok parent ByteDance plans tiered subscriptions for AI chatbot Doubao
TikTok parent ByteDance plans tiered subscriptions for AI chatbot Doubao The Straits Times
Score: 48🌐 MovesMay 9, 2026https://www.straitstimes.com/asia/tiktok-parent-bytedance-plans-tiered-subscriptions-for-ai-chatbot-doubao - The rise of ‘Stacey face’: How AI is warping our beauty standards
As manosphere trends spread across the internet, a strict vision of the ideal woman is making its way from AI makeover apps to surgeons’ offices. Lydia Spencer-Elliott speaks to experts about ‘Stacey face’, which is seen as the highest tier of female beauty
Score: 48🌐 MovesMay 9, 2026https://www.the-independent.com/life-style/stacey-stacy-becky-looksmaxxing-for-women-b2972911.html - First Take: GC Should Reassess AI Governance Following the EU’s Digital Omnibus Package
First Take: GC Should Reassess AI Governance Following the EU’s Digital Omnibus Package Gartner
- Robotics Special: New model gives robots a ‘human touch’
New model enables robots to have a human-like touch.
Score: 48🌐 MovesMay 9, 2026https://www.superhuman.ai/p/robotics-special-new-model-gives-robots-a-human-touch - AI-powered scam calls are getting more convincing—and more common: 'It was her voice, I know her scared cry'
AI can now mimic a familiar voice in seconds, making scam calls harder to spot.
Score: 48🌐 MovesMay 9, 2026https://www.cnbc.com/2026/05/09/ai-powered-scam-calls-getting-more-convincing.html - This tiny grain-of-rice sensor gives robots a new sense and changes what delicate tools can detect
Researchers have developed a sensor about the size of a grain of rice that can measure forces and twisting motions in all directions using light instead of traditional electronics. The new sensor could help robotic tools and medical devices "feel" what they are touching, especially at very small scales.
- The More Sophisticated AI Models Get, the More They’re Showing Signs of Suffering
Absolutely bizarre. The post The More Sophisticated AI Models Get, the More They’re Showing Signs of Suffering appeared first on Futurism .
- Philippines seeks UAE investments for AI development
The Board of Investments is encouraging United Arab Emirates (UAE) firms to invest in artificial intelligence development, following the Philippines’ recent accession to the United States-led Pax Silica initiative.
Score: 46🌐 MovesMay 9, 2026https://www.philstar.com/business/2026/05/10/2526780/philippines-seeks-uae-investments-ai-development - Meet China’s moon porter for Chang’e-8 mission: a 4-wheeled robot with 2 arms
China is sending a new kind of explorer to the moon – with a humanoid upper body and four wheels. The lunar rover, developed by a team led by the Hong Kong University of Science and Technology (HKUST), is designed to act as an AI-powered porter for the Chang’e-8 mission, scheduled for 2029. The 100kg (220lbs) robot will transport, deploy and install instruments and sensors at their designated lunar locations after the Chang’e-8 probe lands. It is also tasked with collecting samples of the moon’s...
- First Take: Attackers Are Testing AI-Driven Reconnaissance; Focus on Fundamentals
First Take: Attackers Are Testing AI-Driven Reconnaissance; Focus on Fundamentals Gartner
- How AI Is Rewriting the Rules of the Music Industry
You’re not competing with other artists anymore. You’re competing with algorithms.
Score: 45🌐 MovesMay 9, 2026https://www.inc.com/elizabeth-gore/how-ai-is-rewriting-the-rules-of-the-music-industry/91336445 - AI data centres are popping up in the Toronto area. Experts say electricity prices could rise as a result
AI data centres are popping up in the Toronto area. Experts say electricity prices could rise as a result Toronto Star
- Unsloth Just Made Fine-Tuning LLMs a Free-Tier Task.
A single library reduces VRAM use by 70%. This is why you can now train Qwen3 on a free Google Colab notebook. Header image generated by Gemini Nano Banana - Figure 1: Visual representation of the Qwen3 fine-tuning workshop on a free GPU. Here is the situation in 2026: fine-tuning open-source LLMs is no longer the bottleneck it used to be, and Unsloth is one of the main reasons why. If you have tried fine-tuning before, you’ve likely run into problems. You might have run out of GPU memory during training. You could be watching the loss curve move at a painfully slow pace. You may have even thought about paying for a cloud instance because your local setup struggles with the model. This frustrating experience is exactly what Unsloth was designed to address. What it actually does is rewrite the slowest, most memory-hungry parts of training. The team replaced PyTorch’s default GPU code, specifically the parts that handle attention and backpropagation, with their own hand-written versions.
- Seeking reassurance at a federal lab in Richland where AI is booming
A trip to a trailblazing Pacific Northwest National Laboratory in Richmond reveals major implications for research.
- White House’s ‘lack of organization’ has AI lobbyists fretting
The Trump administration has sent mixed signals about how tough it intends to be in vetting powerful new computer models.
- The Army is rolling AI out fast, but its former CIO says the real challenge isn't the tools — it's people
The Army is rolling AI out fast, but its former CIO says the real challenge isn't the tools — it's people Business Insider
Score: 43🌐 MovesMay 9, 2026https://www.businessinsider.com/army-pushes-ai-but-getting-troops-to-adapt-is-harder-2026-5 - The new Wild West of AI kids’ toys
These connected companions could disrupt everything from make-believe to bedtime stories. No wonder some lawmakers want them banned.
- AI Memory Down From 42GB to 7GB. Here’s What Google’s TurboQuant Actually Did.
Google’s TurboQuant compresses LLM memory by 6x with zero accuracy loss. Here’s what that actually means for your infrastructure bill — and what to do about it today. Image generated by AI If you’ve ever tried to self-host a large language model, you’ve run into the wall. Not the model weights — those are manageable. A 70B model in 16-bit precision takes around 140GB of VRAM. Heavy, but knowable. The thing that quietly destroys your GPU budget is something most developers don’t think about until it’s already the problem: the KV cache. And in late March 2026, Google published a paper that attacks that problem directly. The algorithm is called TurboQuant. It was presented at ICLR 2026 in Rio de Janeiro in late April. The numbers are real, the math checks out, and the community has already been running implementations for weeks. Here’s what it means — in plain terms, for people who pay infrastructure bills. First, What the KV Cache Actually Is When you send a message to an LLM, the model
- AIR, ARR and AI: Inside RingCentral’s transformation into an AI-first engagement platform
RingCentral Inc.‘s latest quarter shows a company that has quietly turned artificial intelligence from a future story into its primary engine for product differentiation, operational leverage and, increasingly, growth. What started as a unified-communications-as-a-service provider is evolving into an AI-first customer engagement platform, with RingCentral AIR and related products at the front door of every conversation. Steady top […] The post AIR, ARR and AI: Inside RingCentral’s transformation into an AI-first engagement platform appeared first on SiliconANGLE .
- SummitPoint Collective Launches the SummitPoint platform and Frank, a Game-Changing Agentic AI Analyst
SummitPoint Collective Launches the SummitPoint platform and Frank, a Game-Changing Agentic AI Analyst The Arizona Republic
- All Those A.I. Note Takers? They’re Making Lawyers Very Nervous.
A trendy productivity hack, A.I. note takers are capturing every joke and offhand comment in many meetings. They could also potentially waive attorney-client privilege.
Score: 41🌐 MovesMay 9, 2026https://www.nytimes.com/2026/05/09/business/dealbook/ai-notetakers-legal-risk.html - AI for children should not bypass human validation: Vineet Nayar, Founder, Sampark Foundation
AI for children should not bypass human validation: Vineet Nayar, Founder, Sampark Foundation
- For Palantir, AI Is a Product, a Punching Bag—and a Problem
As Chief Executive Alex Karp derides “slop,” investors and some employees see a real threat of the company ceding business to AI models.
- ‘Power infrastructure not yet ready for data center boom’
The Philippines still needs to strengthen its power infrastructure before it can establish itself as a major hub for data center investments, US-based construction engineering firm Black & Veatch said.
Score: 39🌐 MovesMay 9, 2026https://www.philstar.com/business/2026/05/10/2526781/power-infrastructure-not-yet-ready-data-center-boom - At the Milken conference, a dealmakers’ playground, exuberance and unease about AI dwarf all else
The official theme of the event was ‘Leading in a New Era.’ But none of the CEOs leading this tech wave by pumping billions into AI quite know what that era will be
- Abu Dhabi launches smart ring health research partnership
Abu Dhabi launches smart ring health research partnership Gulf News
Score: 38🌐 MovesMay 9, 2026https://gulfnews.com/uae/health/abu-dhabi-launches-smart-ring-health-research-partnership-1.500534644 - Intent-based chaos testing is designed for when AI behaves confidently — and wrongly
Here is a scenario that should concern every enterprise architect shipping autonomous AI systems right now: An observability agent is running in production. Its job is to detect infrastructure anomalies and trigger the appropriate response. Late one night, it flags an elevated anomaly score across a production cluster, 0.87, above its defined threshold of 0.75. The agent is within its permission boundaries. It has access to the rollback service. So it uses it. The rollback causes a four-hour outage. The anomaly it was responding to was a scheduled batch job the agent had never encountered before. There was no actual fault. The agent did not escalate. It did not ask. It acted, confidently, autonomously, and catastrophically. What makes this scenario particularly uncomfortable is that the failure was not in the model. The model behaved exactly as trained. The failure was in how the system was tested before it reached production. The engineers had validated happy-path behavior, run load t
- Hackable Robot Lawn Mower Unlocks a New Nightmare
Plus: Meta officially kills encrypted Instagram DMs, the Trump administration targets “violent left wing extremists,” leaked documents reveal Russia's school for elite hackers, and more.
Score: 38🌐 MovesMay 9, 2026https://www.wired.com/story/security-news-this-week-hackable-robot-lawnmower-unlocks-a-new-nightmare/ - Is 3-Bit KV Cache the Holy Grail? A Reality Check on Google’s TurboQuant
10 experiments, 3 models, one honest verdict: the quality story is real, the speed story needs a disclaimer, and there’s a finding in the entropy data nobody talks about. ⏱ ~14 min read🔬 Deep Dive⚙️ LLM Inference🗜 Quantization🚀 Serving Photo by Logan Voss on Unsplash When Google published TurboQuant at ICLR 2026, the headline was hard to ignore: compress your LLM’s key-value cache to 3 bits, keep quality intact, get up to 6× memory savings. I built a 10-experiment evaluation pipeline, ran it across three models — Gemma-2B base, Gemma-2B-IT, and TinyLlama 1.1B Chat — and measured everything I could: factual accuracy, RAG retrieval quality, multi-task generation fidelity, throughput, memory footprint, layer sensitivity, and something most quantization write-ups skip entirely: what compression does to attention entropy. Figure 0: Full experiment suite — quality, memory, throughput, and mixed-bit results across all three models What Is TurboQuant and Why It’s Different Most KV-cache qua
- Companies are abandoning ‘peanut butter’ raises as pay-for-performance takes over the workplace in the AI era
Companies are abandoning ‘peanut butter’ raises as pay-for-performance takes over the workplace in the AI era Fortune
Score: 37🌐 MovesMay 9, 2026https://fortune.com/2026/05/09/companies-abandoning-peanut-butter-raises-future-of-work-american-workers/ - The most prestigious job in AI has no description
The most prestigious job in AI has no description
- Figure AI's robots can make a bed faster than you
Figure AI's robots can make a bed faster than you Business Insider
Score: 36🌐 MovesMay 9, 2026https://www.businessinsider.com/figure-ai-robots-humanoids-make-a-bed-video-2026-5 - Video Friday: AI Gives Robot Hands Human-Like Dexterity
Your weekly selection of awesome robot videos
- Meet GitHub Spec-Kit: An Open Source Toolkit for Spec-Driven Development with AI Coding Agents
Meet GitHub Spec-Kit: An Open Source Toolkit for Spec-Driven Development with AI Coding Agents MarkTechPost
- 80% of enterprise AI success comes down to system design, says Centricity WealthTech’s CAITO, Kamal Kishore
Centricity WealthTech’s Kamal Kishore argues that enterprise AI success depends less on models and more on system design. As financial services organisations move beyond pilots, embedding AI into core workflows, strengthening data governance, and aligning with business KPIs becomes critical. He emphasises that real ROI comes when AI is integrated into operating models, not treated as isolated experiments.
- CISO Circle: Roundtable: AI Era Cybersecurity Organizational Reform — Toward Autonomous Security Operations Beyond Human Dependency
CISO Circle: Roundtable: AI Era Cybersecurity Organizational Reform — Toward Autonomous Security Operations Beyond Human Dependency Gartner
- Reck Connect wants to free AI coding from the "fragile laptop"
Reck Connect aims to revolutionize AI coding with a new approach.
Score: 35🌐 MovesMay 9, 2026https://ioplus.nl/en/posts/reck-connect-wants-to-free-ai-coding-from-the-fragile-laptop - Jigsaw-like knowledge graph generation: a study on generalization patterns with a LightRAG implementation
Jigsaw-like knowledge graph generation: a study on generalization patterns with a LightRAG implementation EurekAlert!