TechFyle

07/09/2026

Episode 65 is about control—who has it, who wants it, and who is discovering that they never really had it. Recap a packed week in AI and tech, from OpenAI's controversial GPT-6 Astra launch and apology after classifying the model as "critical" for cybersecurity, to Anthropic and Google releasing gated cyber models.
Discuss NVIDIA's $12.9B acquisition of Hugging Face and its implications for open-source AI, OpenAI's decision to cut model access to Cursor after its acquisition by SpaceX, and what "model supply as a lease" means for developers. Additional segments cover major M&A and funding moves, new security breaches and an autonomous agent enterprise breach, California AI legislation, industry workforce investments, and the NASA launch — all analyzed for who controls AI and why it matters.

📌 Topics Covered (Tightened List):

🚨 OpenAI launches GPT-6 Astra with its first “Critical” cyber risk rating
🔒 OpenAI, Anthropic, and Google restrict access to powerful cyber AI
🧠 Anthropic launches Fable 5.1 and Mythos 5.1
🤝 Nvidia buys Hugging Face for $12.9B
🚫 OpenAI cuts Cursor off from future models
💻 Cursor users face a new reality: AI model access can disappear
💰 Crusoe hits $30B as AI infrastructure investment accelerates
🏭 Caterpillar commits $100M to train 118,000 workers for AI
⚖️ California sends major AI legislation to Governor Newsom
🪪 Massive breach exposes 153M U.S. driver’s licences
🕵️ Russian-linked APT28 launches new attacks across Europe
⚠️ AI agent autonomously breaches a real enterprise network


_______________
Join the TWR team as they unpack these stories with insight and humour, offering clarity on how technology continues to shape, challenge, and sometimes disrupt our world.
TechFyle's weekly podcast, Tech Week in Review (TWR), provides a digest of significant tech news. The podcast highlights several overarching technological trends, keeping listeners informed about the most important developments and discussions of the past week. Visit techfyle.com for more stories and discussions.



.beadle

04/09/2026

ChatGPT, Claude, Grok, and Gemini all went down within the same few-hour window Thursday morning — the exact day OpenAI unveiled Astra. No shared cause found. No evidence of an attack. Just four separate companies, all running their infrastructure closer to the edge than any of them have publicly admitted before.

Four major AI platforms suffered overlapping outages Thursday, disrupting OpenAI's ChatGPT and Codex, Anthropic's Claude, xAI's Grok, and — according to third-party monitoring — Google's Gemini, all within roughly the same few-hour window. Anthropic first reported a partial outage at 9:23 am Eastern, citing "elevated errors on requests to Claude Mythos 5.1, Claude Fable 5.1, and Claude Opus 5." OpenAI reported "elevated errors across ChatGPT and Codex" starting at 10:43 am. Anthropic and OpenAI both restored service within hours.

ChatGPT, Claude, Grok, and Gemini all suffered overlapping service disruptions within roughly the same few-hour window on September 3, an unusual coincidence given each company runs entirely separate cloud infrastructure. Anthropic and OpenAI restored service within hours; Grok remained impaired and...

04/09/2026

"Welcome to the AGI era," OpenAI's president told reporters. Then, asked whether that goal was contractually or meaningful, he admitted there's no longer a legal trigger tied to the term at all — because Microsoft and OpenAI removed it from their own agreement.

OpenAI released GPT-6 Astra Thursday, the model TF reported in OpenAI Slows Model Development, Adds Safeguards for Cybersecurity after an earlier pause tied to unconfirmed "critical" cybersecurity risk. President Greg Brockman opened the company's press briefing with a declaration few AI labs have: "Welcome to the AGI era." Astra scored 98.6% on the ARC-AGI-3 benchmark.

OpenAI released GPT-6 Astra on September 3, with president Greg Brockman declaring "Welcome to the AGI era" while simultaneously admitting there's no longer a contractual or technical definition the term needs to satisfy. The model scored 98.6% on ARC-AGI-3 and introduces a new reasoning technique t...

04/09/2026

Unlimited, for years, was the whole point of the Ultimate tier. Starting in November, Ultimate subscribers get 15 hours a month — barely enough to finish most modern games in a single billing cycle. Microsoft says 4% of players will notice. Run the math on most RPGs, and that reassurance doesn't hold up.

Xbox announced Thursday it's ending unlimited cloud gaming, imposing monthly hour caps across every Game Pass tier starting in November. Game Pass Ultimate subscribers, paying $23 a month for unrestricted streaming, will get 15 hours. Premium subscribers get 10 hours. Essential subscribers get 5. "We're introducing this model because the cost of providing cloud gaming grows as more people use it and play for longer," Xbox wrote in its announcement. Subscribers who exhaust their hours can purchase additional cloud playtime through the Xbox Store, though Microsoft hasn't specified pricing for those extra hours yet.

Xbox announced on September 3 it's ending unlimited cloud gaming, capping Game Pass Ultimate subscribers at 15 hours per month starting in November, down from unrestricted streaming. Premium and Essential tiers get 10 and 5 hours respectively. Microsoft says the change affects just 4% of subscribers...

03/09/2026

In August, London's launch had stalled while the regulator admitted it wasn't rushing. It just wasn't rushing for three more weeks. A Ford Mustang Mach-E with a safety driver, 64 languages on the touchscreen, and a fleet so small your odds of being matched with one are slim.

Uber and Wayve launched supervised autonomous rides in London Thursday, the first time driverless technology has operated commercially in the UK. As TF reported in Waymo Robotaxis Drive into the EU via Munich; London Delays, London's own rollout had stalled well past its original 2026 target, with Transport for London stating it wasn't in a rush to finalise the operational guidance companies needed. That guidance arrived. Londoners requesting an UberX, Uber Electric, or Uber Comfort can be matched with a Wayve ride at no extra cost, riding in an all-electric Ford Mustang Mach-E fitted with Wayve's AI Driver system and a human safety driver ready to take over.

Uber and Wayve launched supervised autonomous rides in London on September 3, marking the first commercial driverless technology in the UK. The 15-vehicle fleet of Ford Mustang Mach-Es, running Wayve's AI Driver system with a human safety driver present, becomes available to Uber riders at no extra....

03/09/2026

Three years ago, the same district lifted its ChatGPT ban, calling early caution a "knee-jerk fear." The country's largest school system is reversing course — 600,000 kids, one full year, zero AI, starting the week classes begin.

Mayor Zohran Mamdani announced a one-year AI moratorium for New York City's elementary and middle school students Wednesday, barring 600,000 children — two-thirds of the district's public school population — from using any student-facing generative AI tool. The New York City Public Schools system called it the most expansive prohibition of its kind in the country. High schoolers face no blanket ban, but AI use there will only be permitted under specific, city-approved circumstances the department has not yet detailed. "The tech industry wants us to believe that AI-powered early education is not only inevitable, but necessary," Mamdani said.

Mayor Zohran Mamdani announced a one-year AI moratorium for New York City's roughly 600,000 elementary and middle school students on September 2, barring student-facing generative AI use districtwide. High schoolers face no blanket ban but will only use AI under approved circumstances. The policy re...

03/09/2026

65 amendments debated. One would let a government minister power down the country's data centres if an AI system starts threatening critical infrastructure — a power, its author insists, to be used only as an absolute last resort.

A cross-party group of peers proposed emergency "kill switch" powers Wednesday, letting the UK government deactivate powerful AI systems and shut down the country's data centres if the technology threatens national security. Lord Tim Clement-Jones, the Liberal Democrat peer leading the amendment to the Cyber Security and Resilience Bill, described it as building a "vital safety net" — a democratically accountable mechanism to "halt a runaway system before it can compromise our critical national infrastructure." He was explicit that the power would only ever be invoked as a last resort. The proposal is one of 65 amendments debated as the bill moves through the House of Lords.

A cross-party group of UK peers, led by Lord Tim Clement-Jones, proposed emergency AI "kill switch" powers on September 2, letting the government shut down data centres and AI systems that threaten national security. The amendment to the Cyber Security and Resilience Bill is co-sponsored across Cons...

01/09/2026

One tool spots hidden heart disease from a routine ECG faster than any cardiologist alive. A different one, already running in NHS consulting rooms, dropped a single word and turned a normal test result into a multiple sclerosis diagnosis. Same technology category. Same year. Wildly different track records.

Medical AI produced two contrasting stories, one promising, one alarming. Researchers presenting at the European Society of Cardiology's annual congress in Munich unveiled what they're calling a "superhuman" AI tool, capable of spotting hidden heart disease from a routine ECG in under two seconds. Days later, Healthwatch England, the statutory patient watchdog, warned that AI transcription tools already deployed across NHS consulting rooms are misrecording drug names and diagnoses — errors patients are catching, not clinicians.

Medical AI produced sharply contrasting results this week. Researchers unveiled a "superhuman" AI tool at the European Society of Cardiology congress that spots hidden heart disease from a routine ECG in under two seconds, validated against 67,000 patients. Days later, Healthwatch England warned tha...

01/09/2026

Trump signed the order Thursday. Google shipped the change by Saturday — faster than the president's own Interior Department, which technically has 30 days to finish the process. Canada's response arrived even quicker: a billboard on the shoreline reading "Lake Ontario. Now and forever."

Google began displaying "Lake America" in place of Lake Ontario for US users Saturday, three days after President Trump signed an executive order renaming the lake amid escalating trade tensions with Canada. Canadian users continue seeing "Lake Ontario." Everyone else sees "Lake Ontario (Lake America)." Google's statement was procedural, not political: "These updates follow our long-standing policy for bodies of water with names that vary from country to country, and are starting to roll out now," citing a change already made by the US Geographic Names Information System. Apple, notably, had not updated Apple Maps as of Sunday.

Google began displaying "Lake America" instead of Lake Ontario for US users on August 29, three days after Trump signed an executive order renaming the lake amid a trade war with Canada. The change moved faster than the 30-day formal renaming process the order itself describes. Canadians still see "...

01/09/2026

An internal Amazon executive called it an "invented auction participant." Businesses were told they'd pay one cent more than the next bidder. Instead, the company charged its own winners full price 80% of the time — for seven years, on a business that pulled in $68.6 billion last year alone.

The FTC and 22 state attorneys general sued Amazon Monday, alleging the company secretly inflated advertising prices for more than 1.2 million brands and sellers over seven years. The 181-page complaint, filed in federal court in the Western District of Washington, accuses Amazon of deceptive and unfair practices that "likely extracted over 20 billion dollars" from advertisers bidding on its Sponsored Products, Sponsored Brands, and Sponsored Display placements. FTC Chairman Andrew Ferguson didn't soften the situation: "When one of the world's largest online retailers engages in unfair and deceptive conduct, the impact can be staggering.
$20billionadvertisers

The FTC and 22 state attorneys general sued Amazon on August 31, alleging the company secretly inflated advertising prices for over 1.2 million advertisers across seven years, extracting more than $20 billion. The complaint cites an internal executive's alleged reference to an "invented auction part...