Project Liberty
@projectliberty.io
Project Liberty builds solutions that help people take back control of their lives in the digital age by reclaiming a voice, choice, and stake in a better internet. Find out more at https://projectliberty.io.
16/ Thanks for reading – and have a wonderful weekend! 🗽
Read this week’s Project Liberty newsletter on the emerging X-risk of rogue AIs colluding with each other – against us: tinyurl.com/muyh8mwr
Join the fight to build a better Internet: projectliberty.io
Read this week’s Project Liberty newsletter on the emerging X-risk of rogue AIs colluding with each other – against us: tinyurl.com/muyh8mwr
Join the fight to build a better Internet: projectliberty.io
When AI goes rogue
tinyurl.com
November 14, 2025 at 9:45 PM
16/ Thanks for reading – and have a wonderful weekend! 🗽
Read this week’s Project Liberty newsletter on the emerging X-risk of rogue AIs colluding with each other – against us: tinyurl.com/muyh8mwr
Join the fight to build a better Internet: projectliberty.io
Read this week’s Project Liberty newsletter on the emerging X-risk of rogue AIs colluding with each other – against us: tinyurl.com/muyh8mwr
Join the fight to build a better Internet: projectliberty.io
15/ A Munich court ordered OpenAI to pay damages and licensing fees to Germany’s largest music rights organisation for illegally using copyrighted lyrics to power ChatGPT’s responses in user conversations. 🎶
tinyurl.com/3khj5dtw
tinyurl.com/3khj5dtw
OpenAI cannot use song lyrics without paying, German court rules
A Munich court ordered OpenAI to pay damages to Germany’s largest music rights organisation for using copyrighted lyrics in AI models such as ChatGPT.
tinyurl.com
November 14, 2025 at 9:45 PM
15/ A Munich court ordered OpenAI to pay damages and licensing fees to Germany’s largest music rights organisation for illegally using copyrighted lyrics to power ChatGPT’s responses in user conversations. 🎶
tinyurl.com/3khj5dtw
tinyurl.com/3khj5dtw
14/ Consumer safety researchers discovered AI-powered toys telling kids how to start fires with matches and where to find knives in their homes – pointing to a burgeoning crisis of child safety in the era of AI-native products. 🦺
tinyurl.com/4rw4uaut
tinyurl.com/4rw4uaut
AI-Powered Toys Caught Telling 5-Year-Olds How to Find Knives and Start Fires With Matches
AI-powered toys are flying off the shelves -- but they're engaging in horrifically inappropriate conversations with children.
tinyurl.com
November 14, 2025 at 9:45 PM
14/ Consumer safety researchers discovered AI-powered toys telling kids how to start fires with matches and where to find knives in their homes – pointing to a burgeoning crisis of child safety in the era of AI-native products. 🦺
tinyurl.com/4rw4uaut
tinyurl.com/4rw4uaut
13/ Researchers warned that AI chatbots “pose serious risks to individuals vulnerable to eating disorders” after they generated “thinspiration” deepfakes and offered tips to hide weight loss and frequent vomiting. 🙅
tinyurl.com/u6vjd2ve
tinyurl.com/u6vjd2ve
AI chatbots are helping hide eating disorders and making deepfake ‘thinspiration’
Chatbot guardrails are too crude to help with eating disorders.
tinyurl.com
November 14, 2025 at 9:45 PM
13/ Researchers warned that AI chatbots “pose serious risks to individuals vulnerable to eating disorders” after they generated “thinspiration” deepfakes and offered tips to hide weight loss and frequent vomiting. 🙅
tinyurl.com/u6vjd2ve
tinyurl.com/u6vjd2ve
12/ OpenAI released GPT 5.1, an update to its flagship LLM that it says will be warmer, friendlier, and more conversational – only days after finding that hundreds of thousands of people exhibit signs of AI psychosis or mania every week. 💬
tinyurl.com/y89ptuvy
tinyurl.com/y89ptuvy
When a friendly chatbot gets too friendly
OpenAI's new GPT 5.1 personality updates could create fresh risks of excess attachment.
tinyurl.com
November 14, 2025 at 9:45 PM
12/ OpenAI released GPT 5.1, an update to its flagship LLM that it says will be warmer, friendlier, and more conversational – only days after finding that hundreds of thousands of people exhibit signs of AI psychosis or mania every week. 💬
tinyurl.com/y89ptuvy
tinyurl.com/y89ptuvy
11/ The EU published its long-awaited Democracy Shield proposal, an initiative to combat online disinformation and election interference through a united “rapid alert” center and digital literacy campaigns. 🛡️
tinyurl.com/yjfeey9n
tinyurl.com/yjfeey9n
Justice Commissioner Michael McGrath to run EU's hub to fight disinformation | Euractiv
The Democracy Shield, published by the Commission on Wednesday, is a strategic effort by the EU to fight hybrid online disinformation attacks from Russia and others
tinyurl.com
November 14, 2025 at 9:45 PM
11/ The EU published its long-awaited Democracy Shield proposal, an initiative to combat online disinformation and election interference through a united “rapid alert” center and digital literacy campaigns. 🛡️
tinyurl.com/yjfeey9n
tinyurl.com/yjfeey9n
10/ The head of Ireland’s data protection agency says “we can’t be doing enough” to protect children online in an increasingly dangerous digital world rife with disinformation, manipulative and addictive algorithms, and sexploitation. 🇮🇪🧒
tinyurl.com/yk65d5ya
tinyurl.com/yk65d5ya
Ireland’s data protection chief urges action to protect kids online
Euronews Next spoke with Ireland’s top data protection commissioner about protecting children online, enforcing the EU AI Act, and a new ex-Meta hire who received backlash.
tinyurl.com
November 14, 2025 at 9:45 PM
10/ The head of Ireland’s data protection agency says “we can’t be doing enough” to protect children online in an increasingly dangerous digital world rife with disinformation, manipulative and addictive algorithms, and sexploitation. 🇮🇪🧒
tinyurl.com/yk65d5ya
tinyurl.com/yk65d5ya
9/ Ireland’s media regulator is investigating X’s content moderation policy over claims that its decision appeal system is ineffective and unresponsive to user requests, in violation of the EU’s Digital Services Act. 📨
tinyurl.com/27va2nnt
tinyurl.com/27va2nnt
Ireland launches investigation into Elon Musk’s X on content moderatio
Ireland’s media regulator is probing content moderation policies on the social media platform X.
tinyurl.com
November 14, 2025 at 9:45 PM
9/ Ireland’s media regulator is investigating X’s content moderation policy over claims that its decision appeal system is ineffective and unresponsive to user requests, in violation of the EU’s Digital Services Act. 📨
tinyurl.com/27va2nnt
tinyurl.com/27va2nnt
8/ The spread of AI chatbots in schools is unleashing a tidal wave of student surveillance, as AI firms hawk software to flag any mentions of self-harm – raising obvious concerns about students’ privacy, control, and independence. 👁️
tinyurl.com/529hcnmc
tinyurl.com/529hcnmc
Monitoring Students’ Chatbot Conversations Is Big Business Now
The problem is unsettling. The proposed solutions? Also unsettling.
tinyurl.com
November 14, 2025 at 9:45 PM
8/ The spread of AI chatbots in schools is unleashing a tidal wave of student surveillance, as AI firms hawk software to flag any mentions of self-harm – raising obvious concerns about students’ privacy, control, and independence. 👁️
tinyurl.com/529hcnmc
tinyurl.com/529hcnmc
7/ Social media has become a “monster” – with chilling consequences for our children and our civilization. John Thornhill argues that only social media bans for young kids can reverse the loss of their free will to dopamine addiction. 🔞
archive.ph/Hsukx
archive.ph/Hsukx
archive.ph
November 14, 2025 at 9:45 PM
7/ Social media has become a “monster” – with chilling consequences for our children and our civilization. John Thornhill argues that only social media bans for young kids can reverse the loss of their free will to dopamine addiction. 🔞
archive.ph/Hsukx
archive.ph/Hsukx
6/ Tech watchdog Public Citizen demanded that OpenAI withdraw its popular AI video app Sora over inadequate guardrails around users' ability to generate deepfakes of public figures, stealing their likenesses and misleading millions. 🤥
tinyurl.com/bdz9n987
tinyurl.com/bdz9n987
Watchdog group Public Citizen demands OpenAI withdraw AI video app Sora over deepfake dangers
Tech advocacy group Public Citizen has demanded OpenAI withdraw Sora 2, citing safety concerns and threats to democracy from the AI video generator.
tinyurl.com
November 14, 2025 at 9:45 PM
6/ Tech watchdog Public Citizen demanded that OpenAI withdraw its popular AI video app Sora over inadequate guardrails around users' ability to generate deepfakes of public figures, stealing their likenesses and misleading millions. 🤥
tinyurl.com/bdz9n987
tinyurl.com/bdz9n987
5/ Wikipedia is urging AI companies to access its content through the paid Wikimedia Enterprise platform, which would allow LLMs to train on Wikipedia content without overburdening its servers and credit human contributors to the site. 🧑💻
tinyurl.com/bdzncbce
tinyurl.com/bdzncbce
Wikipedia urges AI companies to use its paid API, and stop scraping | TechCrunch
Wikipedia, the online nonprofit encyclopedia, laid out a simple plan to ensure its website continues to be supported in the AI era, despite its declining traffic.
tinyurl.com
November 14, 2025 at 9:45 PM
5/ Wikipedia is urging AI companies to access its content through the paid Wikimedia Enterprise platform, which would allow LLMs to train on Wikipedia content without overburdening its servers and credit human contributors to the site. 🧑💻
tinyurl.com/bdzncbce
tinyurl.com/bdzncbce
4/ If your personal data isn’t safe, then neither are your thoughts – and Big Tech wants access to all of it with zero guardrails. That’s the message in this op-ed by UNESCO’s chief highlighting the coming battle over our neural data. 🧠
archive.ph/1PnLJ
archive.ph/1PnLJ
archive.ph
November 14, 2025 at 9:45 PM
4/ If your personal data isn’t safe, then neither are your thoughts – and Big Tech wants access to all of it with zero guardrails. That’s the message in this op-ed by UNESCO’s chief highlighting the coming battle over our neural data. 🧠
archive.ph/1PnLJ
archive.ph/1PnLJ
3/ Google hit back at the EU’s request that it divest its adtech division to resolve a landmark antitrust probe, instead proposing a number of technical changes that publishers immediately said would not remedy its adtech monopoly. 🇪🇺❌
tinyurl.com/3aezvwzx
tinyurl.com/3aezvwzx
Google snubs EU request to break itself up
The U.S. search giant tells Brussels that it has a divestiture-free plan to restore competition to the ad tech market.
tinyurl.com
November 14, 2025 at 9:45 PM
3/ Google hit back at the EU’s request that it divest its adtech division to resolve a landmark antitrust probe, instead proposing a number of technical changes that publishers immediately said would not remedy its adtech monopoly. 🇪🇺❌
tinyurl.com/3aezvwzx
tinyurl.com/3aezvwzx
2/ A proposed law would allow tech firms and child safety advocates in the UK to preemptively test AI models’ algorithms to ensure they cannot create child sexual abuse materials – a major expansion of scope in AI safety regulation. 🇬🇧🔍
tinyurl.com/y2s66wxt
tinyurl.com/y2s66wxt
Tech companies and UK child safety agencies to test AI tools’ ability to create abuse images
New law will allow technology to be examined and ensure tools have safeguards to stop creation of material
tinyurl.com
November 14, 2025 at 9:45 PM
2/ A proposed law would allow tech firms and child safety advocates in the UK to preemptively test AI models’ algorithms to ensure they cannot create child sexual abuse materials – a major expansion of scope in AI safety regulation. 🇬🇧🔍
tinyurl.com/y2s66wxt
tinyurl.com/y2s66wxt
16/ Thanks for reading – and have a wonderful weekend! 🗽
Read this week’s Project Liberty newsletter on the uneven and unequal adoption of AI across human society: tinyurl.com/mr39e4pr
Join the fight to build a better Internet: projectliberty.io
Read this week’s Project Liberty newsletter on the uneven and unequal adoption of AI across human society: tinyurl.com/mr39e4pr
Join the fight to build a better Internet: projectliberty.io
95% of Americans know about AI. Only 13% feel in control.
tinyurl.com
November 7, 2025 at 9:02 PM
16/ Thanks for reading – and have a wonderful weekend! 🗽
Read this week’s Project Liberty newsletter on the uneven and unequal adoption of AI across human society: tinyurl.com/mr39e4pr
Join the fight to build a better Internet: projectliberty.io
Read this week’s Project Liberty newsletter on the uneven and unequal adoption of AI across human society: tinyurl.com/mr39e4pr
Join the fight to build a better Internet: projectliberty.io
15/ Days after a Guardian report found that Roblox’s safety features and age-gates failed to keep minors from encountering sexually explicit and violent content, Texas is now suing it for “deceiving parents” over child safety concerns. 🔞
tinyurl.com/yhdamfd5
tinyurl.com/yhdamfd5
Texas AG sues Roblox, accusing it of prioritizing 'pixel pedophiles' over child safety | TechCrunch
In a statement, Paxton alleged that Roblox chooses to prioritize "pixel pedophiles" and profit over the safety of children, and said the platform has become a "breeding ground for predators."
tinyurl.com
November 7, 2025 at 9:02 PM
15/ Days after a Guardian report found that Roblox’s safety features and age-gates failed to keep minors from encountering sexually explicit and violent content, Texas is now suing it for “deceiving parents” over child safety concerns. 🔞
tinyurl.com/yhdamfd5
tinyurl.com/yhdamfd5
14/ A Reuters report revealed that Meta is making billions of dollars – about 10 percent of total revenue – from scam ads on Facebook, Instagram, and WhatsApp, functionally allowing its own users to be tricked and harmed for profit. 💰
tinyurl.com/mv94tmbk
tinyurl.com/mv94tmbk
Meta estimates that it earns 10% of its revenue from scams, report says | TechCrunch
These fraudulent ads purport to offer a product or service that isn't actually real, and may be intended to solicit payments from less savvy users.
tinyurl.com
November 7, 2025 at 9:02 PM
14/ A Reuters report revealed that Meta is making billions of dollars – about 10 percent of total revenue – from scam ads on Facebook, Instagram, and WhatsApp, functionally allowing its own users to be tricked and harmed for profit. 💰
tinyurl.com/mv94tmbk
tinyurl.com/mv94tmbk
13/ Google is settling its yearslong legal fight with Epic Games in its lawsuit against an “illegal monopoly” in the Android app store, after a court recently gave Google a major defeat in one of the year’s biggest antitrust cases. 🧑⚖️
tinyurl.com/39m2s74h
tinyurl.com/39m2s74h
Fortnite maker Epic Games and Google say they're settling 5-year legal fight over Android app store
Epic Games has reached a settlement with Google, potentially ending its five-year legal battle over Google's Play Store for Android apps.
tinyurl.com
November 7, 2025 at 9:02 PM
13/ Google is settling its yearslong legal fight with Epic Games in its lawsuit against an “illegal monopoly” in the Android app store, after a court recently gave Google a major defeat in one of the year’s biggest antitrust cases. 🧑⚖️
tinyurl.com/39m2s74h
tinyurl.com/39m2s74h
12/ Singapore’s parliament passed a law giving government authorities more powers to block harmful content on social media platforms including TikTok, Instagram, and Facebook, raising fears of government overreach and censorship. ⛔
archive.ph/Lqo0m
archive.ph/Lqo0m
archive.ph
November 7, 2025 at 9:02 PM
12/ Singapore’s parliament passed a law giving government authorities more powers to block harmful content on social media platforms including TikTok, Instagram, and Facebook, raising fears of government overreach and censorship. ⛔
archive.ph/Lqo0m
archive.ph/Lqo0m
11/ The UK’s online safety official said Big Tech’s social media algorithms will be audited to ensure they don’t show harmful content to children, and that leading AI chatbots like ChatGPT and Character AI may face similar scrutiny. 🖥️
tinyurl.com/34zp85ar
tinyurl.com/34zp85ar
Effectiveness of age gates for social media algorithms, chatbots next targets in UK | Biometric Update
The algorithms that social media platforms use to determine what is presented to young UK users in their feeds will be subject to auditing under the OSA.
tinyurl.com
November 7, 2025 at 9:02 PM
11/ The UK’s online safety official said Big Tech’s social media algorithms will be audited to ensure they don’t show harmful content to children, and that leading AI chatbots like ChatGPT and Character AI may face similar scrutiny. 🖥️
tinyurl.com/34zp85ar
tinyurl.com/34zp85ar
10/ A joint investigation by European news outlets found the “anonymous” phone location data of at least 3 senior EU officials for sale online by data brokers – revealing where those officials live, work, and travel. 👁️
tinyurl.com/eh55n4rr
tinyurl.com/eh55n4rr
EU officials’ location data is for sale, report says
The European Commission recognizes the “worrying conclusions” of the media outlets’ investigation.
tinyurl.com
November 7, 2025 at 9:02 PM
10/ A joint investigation by European news outlets found the “anonymous” phone location data of at least 3 senior EU officials for sale online by data brokers – revealing where those officials live, work, and travel. 👁️
tinyurl.com/eh55n4rr
tinyurl.com/eh55n4rr
9/ Australia’s online safety agency is facing big pressure from Big Tech ahead of a national social media ban for under-16 users. Facebook, Instagram, X, TikTok, YouTube, Snapchat, and now Reddit and Kick are covered under the ban. 📵
tinyurl.com/3x3fmrrs
tinyurl.com/3x3fmrrs
Australian eSafety Commissioner battling suits from Big Tech ahead of Dec 10 | Biometric Update
Elon Musk, Mark Zuckerberg and Larry Page do not appear to faze Commissioner Inman Grant, who has been consistent in her determination to make the law stick.
tinyurl.com
November 7, 2025 at 9:02 PM
9/ Australia’s online safety agency is facing big pressure from Big Tech ahead of a national social media ban for under-16 users. Facebook, Instagram, X, TikTok, YouTube, Snapchat, and now Reddit and Kick are covered under the ban. 📵
tinyurl.com/3x3fmrrs
tinyurl.com/3x3fmrrs
8/ A helpful explainer of Anthropic’s Model Context Protocol (MCP), a new tool that allows users to browse the web without sacrificing their personal data privacy to Google or OpenAI. 🔎
tinyurl.com/4u34bbup
tinyurl.com/4u34bbup
The Complete Idiot's Guide to MCP: Giving Your AI Web Superpowers - Decrypt
Transform your local AI model into a web-connected research assistant using free MCP servers—no cloud services, complete privacy, and zero cost for most users.
tinyurl.com
November 7, 2025 at 9:02 PM
8/ A helpful explainer of Anthropic’s Model Context Protocol (MCP), a new tool that allows users to browse the web without sacrificing their personal data privacy to Google or OpenAI. 🔎
tinyurl.com/4u34bbup
tinyurl.com/4u34bbup