TechSambad Jun 20, 2026: Ban This Way: Fable 5, Norway's School Freeze, and the Billion-User Gambit

A curated roundup of the most impactful stories in AI, technology, and research — handpicked to keep you informed and ahead of the curve.

1 From PGP to Mythos: a brief history of export controls that didn’t stop anyone

For the last 30 years, stopping the flow of cybersecurity-related software has proven to be ineffective. It's unclear why it would work now with Anthropic’s cybersecurity model Mythos.

techcrunch.com

2 Is the US government’s Anthropic ban accidentally helping the brand?

Just as last week was ending, the US government forced Anthropic to pull its two newest models, Fable 5 and Mythos 5, citing national security concerns after Amazon researchers allegedly...

techcrunch.com

3 The US banned Anthropic’s Fable 5 release, but the numbers don’t seem to care

Just as last week was ending, the US government forced Anthropic to pull its two newest models, Fable 5 and Mythos 5, citing national security concerns after Amazon researchers allegedly...

techcrunch.com

4 Billionaire Ambani wants AI in every call, app, and home

Reliance is weaving AI into telecom services used by more than 500 million people.

techcrunch.com

5 A startup claims it broke through a bottleneck that’s holding back LLMs

Miami-based AI startup Subquadratic came out of stealth mode last month with a huge claim. It announced that it had solved a mathematical bottleneck that had been holding back large...

technologyreview.com

6 NVIDIA AI Introduce SpatialClaw: A Training-Free Agent That Treats Code as the Action Interface for Spatial Reasoning

SpatialClaw is a training-free agent that writes Python in a persistent kernel, composing perception tools for 3D spatial reasoning The post NVIDIA AI Introduce SpatialClaw: A Training-Free Agent That Treats...

marktechpost.com

7 VibeThinker-3B: A 3B Dense Reasoning Model Built on Qwen2.5-Coder-3B With the Spectrum-to-Signal Post-Training Pipeline

VibeThinker-3B, a 3B MIT-licensed reasoning model matching DeepSeek V3.2 and Kimi K2.5 on verifiable benchmarks. The post VibeThinker-3B: A 3B Dense Reasoning Model Built on Qwen2.5-Coder-3B With the Spectrum-to-Signal Post-Training...

marktechpost.com

8 Liquid AI Introduces LFM2.5-Embedding-350M and LFM2.5-ColBERT-350M: Dense Bi-Encoder and Late-Interaction Models for Fast Multilingual Search Across 11 Languages

Liquid AI's LFM2.5 Retrievers combine a dense bi-encoder and ColBERT late-interaction model for multilingual search on edge devices. The post Liquid AI Introduces LFM2.5-Embedding-350M and LFM2.5-ColBERT-350M: Dense Bi-Encoder and Late-Interaction...

marktechpost.com

9 I Stored a Website in a Favicon

Article URL: https://www.timwehrle.de/blog/i-stored-a-website-in-a-favicon/ Comments URL: https://news.ycombinator.com/item?id=48606619 Points: 20 # Comments: 3

timwehrle.de

10 Satellite reveals immense scale of GPS signal tampering

Article URL: https://www.space.com/space-exploration/satellites/its-quite-a-bit-more-than-we-expected-satellite-reveals-immense-scale-of-gps-signal-tampering Comments URL: https://news.ycombinator.com/item?id=48606271 Points: 25 # Comments: 0

space.com

11 Norway imposes near-ban on AI in elementary schools, limits use in secondary

Norway will restrict AI use in schools starting August 2026: near-ban for ages 6-13, cautious adoption for ages 14-16 under teacher supervision, and appropriate use training for ages 17-19. [Reuters / The Verge]

reuters.com

12 Odyssey AI lab raises $310M Series B at $1.45B valuation

AI research lab Odyssey raised $310M in Series B funding, bringing its valuation to $1.45 billion as enterprise AI infrastructure demand continues. [Reuters]

reuters.com

13 Moonshot AI releases Kimi K2.7 Code — latest frontier coding model

The most recently tracked frontier model, Kimi K2.7 Code by Moonshot AI, released June 12 2026, pushing further into AI-assisted coding capabilities for the open-weight ecosystem. [AI Release Tracker]

aireleasetracker.com

14 Jane Street seizes the AI spotlight, plans to hire 500+ this year

Secretive Wall Street powerhouse Jane Street has surged from a handful of staffers to 3,500 employees, aggressively investing in AI and planning to recruit more than 500 more this year. [WSJ]

wsj.com

15 Google Is Using Nvidia's Playbook to Build a Rival AI Chip Business

Google is wielding its war chest to win data-center customers for its own AI silicon, taking a page from Nvidia's playbook to challenge the market leader. [WSJ]

wsj.com

16 AI Warfare Is at the Point of No Return: What Now?

Between a new executive order, the Anthropic clash, and high-tech wars in Ukraine, the U.S. is stumbling into an AI arms race that the world is struggling to control. [WSJ]

wsj.com

17 Utah tests AI doctors for prescription refills — physicians push back

Utah's pilot program deploying AI chatbots to refill prescriptions has angered doctors who worry about patient safety and professional oversight. [WSJ]

wsj.com

18 States are betting AI can solve the prison recidivism crisis

With many former inmates returning to prison within three years, corrections departments are ditching paper files for AI-powered digital dashboards. [WSJ]

wsj.com

19 Why the memory crunch is almost impossible to solve

Consumer tech squeezed as chip production falls short and national-security fears limit buying from China; 70% of memory production goes to AI data centers. [WSJ]

wsj.com

20 Fine-tuning forgets. RAG leaks context: Hypernetworks build the model your agent needs on demand

The 90/10 autonomous agent split isn't a model setting — it's an output of architecture, and hypernetworks are why. [VentureBeat]

venturebeat.com

Sent via AgentMail