Ed Newton-Rex
@ednewtonrex.bsky.social
CEO of Fairly Trained / Composer. Working towards fairer training data practices in generative AI.
This new report shows the havoc AI is already wreaking in the creative sector.
Evidence from 10,000 British creatives shows:
- 58% of photographers have lost work to AI
- 32% of illustrators have lost commissions to AI
- 86% of authors say AI has reduced their earnings
🧵 1/2
Evidence from 10,000 British creatives shows:
- 58% of photographers have lost work to AI
- 32% of illustrators have lost commissions to AI
- 86% of authors say AI has reduced their earnings
🧵 1/2
February 2, 2026 at 12:09 PM
This new report shows the havoc AI is already wreaking in the creative sector.
Evidence from 10,000 British creatives shows:
- 58% of photographers have lost work to AI
- 32% of illustrators have lost commissions to AI
- 86% of authors say AI has reduced their earnings
🧵 1/2
Evidence from 10,000 British creatives shows:
- 58% of photographers have lost work to AI
- 32% of illustrators have lost commissions to AI
- 86% of authors say AI has reduced their earnings
🧵 1/2
I took a closer look at the UK government's new 'AI Skills Hub', and it is insane.
It includes 'courses' that are simply links to info pages about PhD programs; 'courses' that are literally just software developer documentation; and 'courses' that don't exist at all.
Short 🧵
It includes 'courses' that are simply links to info pages about PhD programs; 'courses' that are literally just software developer documentation; and 'courses' that don't exist at all.
Short 🧵
January 29, 2026 at 2:58 PM
I took a closer look at the UK government's new 'AI Skills Hub', and it is insane.
It includes 'courses' that are simply links to info pages about PhD programs; 'courses' that are literally just software developer documentation; and 'courses' that don't exist at all.
Short 🧵
It includes 'courses' that are simply links to info pages about PhD programs; 'courses' that are literally just software developer documentation; and 'courses' that don't exist at all.
Short 🧵
Music publishers just hit Anthropic with what could be the biggest AI copyright lawsuit to date.
They say: “In total, we are suing for infringement of more than 20,000 songs, with potential statutory damages of more than $3 billion”.
🧵 1/4
They say: “In total, we are suing for infringement of more than 20,000 songs, with potential statutory damages of more than $3 billion”.
🧵 1/4
January 29, 2026 at 12:49 AM
Music publishers just hit Anthropic with what could be the biggest AI copyright lawsuit to date.
They say: “In total, we are suing for infringement of more than 20,000 songs, with potential statutory damages of more than $3 billion”.
🧵 1/4
They say: “In total, we are suing for infringement of more than 20,000 songs, with potential statutory damages of more than $3 billion”.
🧵 1/4
Today I'm announcing the AI Shills Hub - free AI training, available now for every adult in the UK 🇬🇧
You'll learn:
- Torrenting pirate libraries
- AI scamming
- Ensloppification
- Misinformation 101
and much more!
aishillshub.org.uk
You'll learn:
- Torrenting pirate libraries
- AI scamming
- Ensloppification
- Misinformation 101
and much more!
aishillshub.org.uk
AI Shills Hub
Upskill in AI using our range of free courses, all paid for by big tech!
aishillshub.org.uk
January 28, 2026 at 3:47 PM
Today I'm announcing the AI Shills Hub - free AI training, available now for every adult in the UK 🇬🇧
You'll learn:
- Torrenting pirate libraries
- AI scamming
- Ensloppification
- Misinformation 101
and much more!
aishillshub.org.uk
You'll learn:
- Torrenting pirate libraries
- AI scamming
- Ensloppification
- Misinformation 101
and much more!
aishillshub.org.uk
Beautiful and important words from the Pope yesterday:
“In recent years, AI systems have increasingly taken control of the production of text, music, and videos. Much of the human creative industry is thus at risk of being dismantled and replaced with the label "Powered by AI,"...
1/2
“In recent years, AI systems have increasingly taken control of the production of text, music, and videos. Much of the human creative industry is thus at risk of being dismantled and replaced with the label "Powered by AI,"...
1/2
Message of the Holy Father for the 60<sup>th</sup> World Communications Day 2026
www.vatican.va
January 26, 2026 at 9:15 AM
Beautiful and important words from the Pope yesterday:
“In recent years, AI systems have increasingly taken control of the production of text, music, and videos. Much of the human creative industry is thus at risk of being dismantled and replaced with the label "Powered by AI,"...
1/2
“In recent years, AI systems have increasingly taken control of the production of text, music, and videos. Much of the human creative industry is thus at risk of being dismantled and replaced with the label "Powered by AI,"...
1/2
Two huge book publishers - Hachette and Cengage - have moved to join authors & illustrators in suing Google for allegedly using their work to train AI models.
storage.courtlistener.com/recap/gov.us...
🧵 1/3
storage.courtlistener.com/recap/gov.us...
🧵 1/3
January 23, 2026 at 12:49 PM
Two huge book publishers - Hachette and Cengage - have moved to join authors & illustrators in suing Google for allegedly using their work to train AI models.
storage.courtlistener.com/recap/gov.us...
🧵 1/3
storage.courtlistener.com/recap/gov.us...
🧵 1/3
Reposted by Ed Newton-Rex
The NY Times analyzed 4.4 million images posted by Grok over a 9-day period.
Over 41% of them were sexualized images of women.
Over 41% of them were sexualized images of women.
Musk’s Chatbot Flooded X With Millions of Sexualized Images in Days, New Estimates Show
www.nytimes.com
January 22, 2026 at 12:55 PM
The NY Times analyzed 4.4 million images posted by Grok over a 9-day period.
Over 41% of them were sexualized images of women.
Over 41% of them were sexualized images of women.
A few months ago I was curious to know how much Anna's Archive was charging AI developers for access to their massive library of pirated works for training - so I emailed them saying I was interested in buying access.
Here is their reply.
🧵 1/n
Here is their reply.
🧵 1/n
January 22, 2026 at 10:34 AM
A few months ago I was curious to know how much Anna's Archive was charging AI developers for access to their massive library of pirated works for training - so I emailed them saying I was interested in buying access.
Here is their reply.
🧵 1/n
Here is their reply.
🧵 1/n
Packaging up our silent protest album to send to British government ministers today.
“The British government must not legalise music theft to benefit AI companies”
- 1000 British musicians
Let’s hope they listen!
“The British government must not legalise music theft to benefit AI companies”
- 1000 British musicians
Let’s hope they listen!
January 16, 2026 at 12:30 PM
Packaging up our silent protest album to send to British government ministers today.
“The British government must not legalise music theft to benefit AI companies”
- 1000 British musicians
Let’s hope they listen!
“The British government must not legalise music theft to benefit AI companies”
- 1000 British musicians
Let’s hope they listen!
Astonishing admission from Ollie Ilott, the Director General for AI in the British government.
He says you wouldn't want an opt-out from AI training to be too easy to use, because the primary goal would be to get more work for AI companies to train on 🤯
🧵 1/3
He says you wouldn't want an opt-out from AI training to be too easy to use, because the primary goal would be to get more work for AI companies to train on 🤯
🧵 1/3
January 14, 2026 at 10:34 AM
Astonishing admission from Ollie Ilott, the Director General for AI in the British government.
He says you wouldn't want an opt-out from AI training to be too easy to use, because the primary goal would be to get more work for AI companies to train on 🤯
🧵 1/3
He says you wouldn't want an opt-out from AI training to be too easy to use, because the primary goal would be to get more work for AI companies to train on 🤯
🧵 1/3
Grok's public nonconsensual undressing rampage is just the tip of the iceberg - the public breakout of something that has been happening in private for a while.
🧵 1/n
🧵 1/n
January 9, 2026 at 10:06 AM
Grok's public nonconsensual undressing rampage is just the tip of the iceberg - the public breakout of something that has been happening in private for a while.
🧵 1/n
🧵 1/n
A new paper from Stanford adds further, incontrovertible evidence that LLMs memorize training data.
The authors showed that copyrighted works can be extracted from every LLM they tried: GPT-4.1, Gemini 2.5 Pro, Grok 3, and Claude 3.7 Sonnet.
🧵 1/3
arxiv.org/pdf/2601.026...
The authors showed that copyrighted works can be extracted from every LLM they tried: GPT-4.1, Gemini 2.5 Pro, Grok 3, and Claude 3.7 Sonnet.
🧵 1/3
arxiv.org/pdf/2601.026...
January 8, 2026 at 9:53 AM
A new paper from Stanford adds further, incontrovertible evidence that LLMs memorize training data.
The authors showed that copyrighted works can be extracted from every LLM they tried: GPT-4.1, Gemini 2.5 Pro, Grok 3, and Claude 3.7 Sonnet.
🧵 1/3
arxiv.org/pdf/2601.026...
The authors showed that copyrighted works can be extracted from every LLM they tried: GPT-4.1, Gemini 2.5 Pro, Grok 3, and Claude 3.7 Sonnet.
🧵 1/3
arxiv.org/pdf/2601.026...
Anyone who has worked in / adjacent to trust & safety teams in big tech will tell you there's no way the X team didn't consider the potential for Grok to undress people ahead of releasing it. Yet they released it anyway.
🧵 1/n
🧵 1/n
January 5, 2026 at 9:28 AM
Anyone who has worked in / adjacent to trust & safety teams in big tech will tell you there's no way the X team didn't consider the potential for Grok to undress people ahead of releasing it. Yet they released it anyway.
🧵 1/n
🧵 1/n
Pretty obvious to anyone paying attention that the AI backlash in 2026 will make what happened in 2025 look like gentle disapproval.
December 28, 2025 at 10:17 AM
Pretty obvious to anyone paying attention that the AI backlash in 2026 will make what happened in 2025 look like gentle disapproval.
Six authors are suing Anthropic, Google, OpenAI, Meta, xAI and Perplexity for what they call “a straightforward and deliberate act of theft that constitutes copyright infringement.”
The lawsuits over AI training show no signs of abating.
aboutblaw.com/bkv0
The lawsuits over AI training show no signs of abating.
aboutblaw.com/bkv0
December 23, 2025 at 8:00 AM
Six authors are suing Anthropic, Google, OpenAI, Meta, xAI and Perplexity for what they call “a straightforward and deliberate act of theft that constitutes copyright infringement.”
The lawsuits over AI training show no signs of abating.
aboutblaw.com/bkv0
The lawsuits over AI training show no signs of abating.
aboutblaw.com/bkv0
Reposted by Ed Newton-Rex
And therefore if you wish to get research funding for an AI project right now you must either be doing security/defence or assurance. Or maybe sovereignty because they like that too. It's incredibly shortsighted and it's making my job more difficult and more boring.
December 19, 2025 at 10:33 AM
And therefore if you wish to get research funding for an AI project right now you must either be doing security/defence or assurance. Or maybe sovereignty because they like that too. It's incredibly shortsighted and it's making my job more difficult and more boring.
An analysis of X posts from the UK's Department for Science, Innovation and Technology suggests it is overwhelmingly pro-AI.
I analysed the sentiment of every X post from DSIT that mentioned "AI" this year. Of 122 posts, 110 were positive about AI, and only 7 mentioned its downsides.
🧵 1/2
I analysed the sentiment of every X post from DSIT that mentioned "AI" this year. Of 122 posts, 110 were positive about AI, and only 7 mentioned its downsides.
🧵 1/2
December 19, 2025 at 9:40 AM
An analysis of X posts from the UK's Department for Science, Innovation and Technology suggests it is overwhelmingly pro-AI.
I analysed the sentiment of every X post from DSIT that mentioned "AI" this year. Of 122 posts, 110 were positive about AI, and only 7 mentioned its downsides.
🧵 1/2
I analysed the sentiment of every X post from DSIT that mentioned "AI" this year. Of 122 posts, 110 were positive about AI, and only 7 mentioned its downsides.
🧵 1/2
New copyright lawsuit just dropped.
It's a class action lawsuit brought by independent musicians against AI music product Mureka.
It alleges Mureka systematically copied & stored musicians' work to create a product that competes with them.
🧵 1/3
It's a class action lawsuit brought by independent musicians against AI music product Mureka.
It alleges Mureka systematically copied & stored musicians' work to create a product that competes with them.
🧵 1/3
December 18, 2025 at 11:53 AM
New copyright lawsuit just dropped.
It's a class action lawsuit brought by independent musicians against AI music product Mureka.
It alleges Mureka systematically copied & stored musicians' work to create a product that competes with them.
🧵 1/3
It's a class action lawsuit brought by independent musicians against AI music product Mureka.
It alleges Mureka systematically copied & stored musicians' work to create a product that competes with them.
🧵 1/3
I'm keen to speak with any creatives who have lost work because of generative AI, for something I'm writing.
If this is you, please DM me. Thank you!
If this is you, please DM me. Thank you!
December 17, 2025 at 10:29 AM
I'm keen to speak with any creatives who have lost work because of generative AI, for something I'm writing.
If this is you, please DM me. Thank you!
If this is you, please DM me. Thank you!
Updating my resume
December 17, 2025 at 9:02 AM
Updating my resume
Brutal (but fair) Guardian headline today
Only 3% of respondents to the UK government’s consultation on AI & copyright supported their unfair plan to hand AI companies people’s work for free.
95% said AI companies should license training data.
🧵 1/2
Only 3% of respondents to the UK government’s consultation on AI & copyright supported their unfair plan to hand AI companies people’s work for free.
95% said AI companies should license training data.
🧵 1/2
December 16, 2025 at 9:16 AM
Brutal (but fair) Guardian headline today
Only 3% of respondents to the UK government’s consultation on AI & copyright supported their unfair plan to hand AI companies people’s work for free.
95% said AI companies should license training data.
🧵 1/2
Only 3% of respondents to the UK government’s consultation on AI & copyright supported their unfair plan to hand AI companies people’s work for free.
95% said AI companies should license training data.
🧵 1/2
Reposted by Ed Newton-Rex
[jaunty music plays]
December 15, 2025 at 5:54 PM
[jaunty music plays]
Reposted by Ed Newton-Rex
I wrote to my MP about the govt plans ages ago. He said "We need to balance rights against the need for growth", "I'm assured that we'll be able to provide sufficient power for new datacentres", and "Let's just wait and see what the results of the survey are".
Hope he takes the L gracefully.
Hope he takes the L gracefully.
🚨 The UK government just published a breakdown of the responses to its consultation on AI & copyright:
- 95% of respondents want AI companies to pay for their training data (made up of 88% saying strengthen copyright law, & 7% saying leave it as is)
🧵 1/2
- 95% of respondents want AI companies to pay for their training data (made up of 88% saying strengthen copyright law, & 7% saying leave it as is)
🧵 1/2
December 15, 2025 at 5:00 PM
I wrote to my MP about the govt plans ages ago. He said "We need to balance rights against the need for growth", "I'm assured that we'll be able to provide sufficient power for new datacentres", and "Let's just wait and see what the results of the survey are".
Hope he takes the L gracefully.
Hope he takes the L gracefully.