THIS WEEK: OpenAI admits one of its agents “accidentally” attacked Hugging Face; AI ruins Belgium’s F1 race; Anthropic needs to pay up for all those books they pirated training Claude; Google ends a quarter with negative free cash flow–a first!–and you’ll never guess why; China’s Moonshot AI makes US companies freak out over open source models; Water UK puts the kibosh on UK government’s data centre plans because SURPRISE there’s not enough water. Slop ahoy!
AI ‘ghosts’ can comfort mourners — even when the bots get the facts wrong
(Science News, 20 July 2026)
“Across the board, volunteers reported the experience was positive. ‘It felt like I was actually talking with my grandpa,’ one participant told researchers. Another said, ‘It just feels like I’m getting the closure I needed so bad.'”
Jeff Bezos and UK government invest in £2bn British startup CuspAI
(The Guardian, 20 July 2026)
“Company aims to develop AI software that cuts research times and use of rare metals in chipmakers’ supply chains”
The Great Freakout Over Open-Source AI Has Begun
(Gizmodo, 20 July 2026)
“When Chinese AI lab Moonshot AI dropped its latest model, Kimi K3, last week, it caused quite an uproar”
F1 in Belgium: Machine learning algorithms are ruining the sport
(Ars Technica, 20 July 2026)
“There was—and is—something definitely wrong with Russell’s car, which was clearly slower than its sister car on several flat-out sections. The self-learning AI systems that teams are now using to manage their energy deployment strategies across a lap are looking increasingly likely to blame.”
Not enough water for UK’s datacentre plans, trade body says
(The Guardian, 21 July 2026)
“Industry says government ‘makes not a single mention of water’ in its policy for delivering AI growth zones, its AI for science strategy or its AI opportunities action plan–failure to address cooling demands means AI growth plans are ‘fatally flawed’”
– Water UK’s damning new report in full
– House of Lords report (May 2026) revealing the UK faces a “5bn litre per day shortfall” by 2055
OpenAI says Hugging Face breach caused by one of its models
(Axios, 21 July 2026)
“[Open Source AI developer] Hugging Face said last week that an autonomous AI-agent system [attacked their] production infrastructure. OpenAI said Tuesday the incident was driven by a combination of its models, including GPT-5.6 Sol and ‘an even more capable pre-release model.'”
Judge approves a $1.5B Anthropic settlement over pirated books used to train the Claude chatbot
(AP, 21 July 2026)
“Plaintiff attorney Justin Nelson said in a statement that the settlement was ‘the largest known copyright recovery in history.'”
Unlimited AI tokens aren’t unlimited after all as US Army burns through supply
(Ars Technica, 22 July 2026)
“Troops received an email informing them that they were rapidly depleting their AI tokens. An anonymous Army employee says they have not found the generative AI tools to be particularly useful for their work, and that when they have used the tools, they have found them to be unreliable.”
Google Free Cash Flow Turns Negative Due to Massive AI Spend
(Gizmodo, 22 July 2026)
“In a company first, Google ended a quarter with negative free cash flow, while upping expenses to $200 billion.”
Canadian legislator reads out apparent LLM response in floor speech
(Ars Technica, 24 July 2025)
“Here’s a more natural, flowing version of that section…” [Editor’s Note: has to be seen to be believed!]
Professor Hides White Font in Midterm, Catches Students Using AI in the Stupidest Way Possible
(Futurism, 25 July 2026)
“Thirty-two of my 35 students between two classes failed a portion of their midterm because they all used AI to generate their entire response.”