Good morning. It’s Thursday, June 11. Today we are covering:
Anthropic reverses course on Claude’s secret AI safeguard policy after backlash, promising more transparency
OpenAI reportedly weighs steep price cuts for ChatGPT as competition with Anthropic intensifies
FBI shuts down 13 Chinese-linked websites that used AI-generated personas to recruit Americans with security clearances
Apple and London’s Metropolitan Police team up to make stolen phones unusable
YouTube rolls out new in-app messaging and group chat features
Let’s dive in
Anthropic Responds to Backlash on Claude’s Secret Sabotage on AI Research
By Max Zeff via Wired
Anthropic reversed a controversial policy that let Claude quietly flag and limit responses to AI research questions it judged risky, after backlash from researchers and developers over the lack of transparency.
Going forward, when Claude restricts or declines a query tied to AI capabilities research, it will say so explicitly, and flagged queries will fall back to Claude Opus 4.8 with a visible explanation of why.
The reversal comes amid growing scrutiny of how AI labs handle dual-use research, and Anthropic said it is committed to keeping its safety measures visible to users and the developer community going forward.
𝕏: Anthropic is walking back Claude Fable 5’s policy to covertly degrade performance for competing AI researchers, after facing fierce backlash. “We’re changing Fable 5’s safeguards for frontier LLM development to make them visible,” Anthropic tells WIRED. “We made the wrong tradeoff and we apologize for not getting the balance right.” - Max Zeff (@ZeffMax)
OpenAI Considers Drastic Price Cuts Anticipating War for Users With Anthropic
By Keach Hagey via The Wall Street Journal
OpenAI is reportedly considering steep price cuts across its ChatGPT subscription tiers as it braces for an intensifying battle for users with Anthropic.
The potential cuts would target both consumer and business plans, aiming to make ChatGPT more competitive on price as rivals like Claude gain ground with developers and enterprises.
[PAYWALL] The move underscores how pricing is becoming a new front in the rivalry between OpenAI and Anthropic as both companies fight for market share.
𝕏: If OpenAI does drop token pricing, this is likely because they’ve heard from customers they can’t adopt AI at volume at the current pricing. Margin is high now for served tokens. They could cut prices by like 60% and still be profitable imo - Max Weinbach (@mweinbach)
The best way to reach new readers is through word of mouth. If you click THIS LINK in your inbox, it’ll create an easy-to-send pre-written email you can just fire off to some friends.
FBI Shutters Covert Chinese Websites Targeting Secrets
By Bill Gertz via The Washington Times
The FBI and Justice Department shut down 13 internet domains used by Chinese actors posing as fake consulting firms to recruit Americans with access to classified and sensitive government information.
The fraudulent companies used AI-generated photos and personas to build credibility, posting job listings on LinkedIn, Upwork, and other sites for roles like “Senior Analyst” to lure security clearance holders into sharing reports via Telegram.
The seizures follow a joint FBI and MI5 warning last week that Chinese intelligence services are using professional networking sites to target defense and foreign policy analysts.
𝕏: “The fake consulting company domains seized by the FBI illustrate the lengths the Chinese government’s intelligence services will go to as they try to use AI-generated content to trick, recruit, or coerce current and former U.S. security clearance holders into sharing sensitive information,” said Assistant Director Roman Rozhavsky of the FBI’s Counterintelligence and Espionage Division. - Bill Bishop (@niubi)
Make Stolen Phones Unusable, Met Police Urges Tech Giants
By Yang Tian, Tony Grew and Liv McMahon via BBC
London’s Metropolitan Police is calling on tech firms to make stolen phones harder to resell, after starting to share data with Apple to track what happens to stolen handsets.
Apple says it has largely closed the “factory reset” loophole that let thieves wipe and resell iPhones, helped by Stolen Device Protection, a security feature now on by default in iOS 26.4.
Phone thefts in London fell 18% year over year, and Commissioner Sir Mark Rowley is pushing for legislation requiring phone companies to publish stolen-device data, with Samsung and Google also making security changes.
𝕏: A positive development. Every tech company should do the same. We need to use every tool possible to stop these criminals. - Matt Vickers MP (@Matt_VickersMP)
YouTube Rolls Out New Messaging and Chat Features
By Abner Li via 9to5Google
YouTube is rolling out a new in-app messaging feature that lets users chat directly with friends without leaving the app.
The chat lives in its own dedicated tab, letting people share videos, Shorts, and channel recommendations in real time with friends and groups.
The rollout is part of Google’s push to make YouTube feel more social, putting it in more direct competition with the built-in messaging features on Instagram and TikTok.
𝕏: Youtube is rolling out direct messages! This will allow you to send and reply to videos without leaving the app! This is accessed via the share menu. Will you be using this feature? - Beezy (@ObeezyLive)
Trending in AI
Anthropic CEO Says Government Must Be Able to Block Dangerous AI
OpenAI Warns Tariffs on China Could Threaten Its AI Data Center Plans
US Seizes 13 Website Domains Tied to Alleged Chinese Intelligence Collection
Nvidia Is Developing an AI Healthcare Model With Startup Abridge
DoorDash Rolls Out AI Ordering Tools With Ask DoorDash Chatbot
Thanks for reading to the bottom and soaking in our Newslit Daily fueled with highlights for your morning.
I hope you found it interesting and, needless to say, if you have any questions or feedback let me know by hitting reply.
Take care and see you tomorrow!
How was today’s email?
