News/AI Safety
Sheryl Crow criticizes Trump era and AI, has a feeling she’s not the only one
Sheryl Crow has released "The New Normal," a new single critiquing the current political climate under President Trump while addressing concerns about artificial intelligence and societal shifts. The song, performed with her touring band the Real Lowdown, reflects the singer's fears that increasingly "bizarre" circumstances might become accepted as ordinary. What she's saying: Crow expressed her concerns about current events in social media posts accompanying the release.• "What's happening all around us is so unbelievably bizarre, that my fear is we will truly begin to feel like this is normal. And that's what terrifies me," she wrote on Instagram. Key...
read Jul 18, 2025Georgia opens AI Innovation Lab for ethical government testing
Georgia's technology department officially opened its Innovation Lab this week, creating a dedicated space for ethical AI experimentation aimed at advancing public service. The facility provides state agencies, cities, counties, and school systems with a hands-on, low-risk environment to test AI use cases before deployment, addressing widespread apprehension about AI adoption through collaboration and transparency. What you should know: The Innovation Lab occupies the entire seventh floor of the Georgia Technology Authority (GTA) headquarters and offers three sandbox environments powered by major cloud providers. The lab runs on Amazon Web Services, Azure, and Google Cloud platforms, pre-loaded with sample data...
read Jul 18, 2025NYPD taps FDNY to bypass facial recognition ban
The NYPD circumvented its facial recognition ban by using the New York Fire Department to identify a pro-Palestinian protester at Columbia University, according to court documents and a recent lawsuit. The arrangement allowed police to access Clearview AI facial recognition technology and DMV records that are typically restricted under NYPD policy, raising significant concerns about surveillance oversight and government accountability. What happened: Fire marshals used Clearview AI to identify Zuhdi Ahmed, a 21-year-old CUNY pre-med student, after he was accused of throwing a rock at a pro-Israeli protester during April 2024 demonstrations at Columbia University. The FDNY searched social media...
read Jul 18, 2025Study reveals 12.8B-image AI dataset contains millions of personal documents
A new study reveals that DataComp CommonPool, one of the largest open-source AI training datasets with 12.8 billion samples, contains millions of images with personally identifiable information including passports, credit cards, birth certificates, and identifiable faces. The findings highlight a fundamental privacy crisis in AI development, as researchers estimate hundreds of millions of personal documents may be embedded in datasets used to train popular image generation models like Stable Diffusion and Midjourney. What you should know: Researchers audited just 0.1% of CommonPool's data and found thousands of validated identity documents and over 800 job application materials linked to real people....
read Jul 18, 2025ChatGPT may watermark AI images for free users only
OpenAI is testing watermarks on AI-generated images in ChatGPT, with evidence found in the Android beta version suggesting the feature may be limited to free users. The move could represent a significant shift in how AI companies monetize their image generation services, potentially forcing users to upgrade to paid plans to avoid watermarked content. What you should know: The watermark feature was discovered through code references and beta testing, though its final implementation remains unclear. Android Authority spotted a "save without watermark" option in the latest ChatGPT beta for Android, originally identified by X user Tibor Blaho who found "image-gen-watermark-for-free"...
read Jul 18, 2025“Nobody could convince me it wasn’t her.” AI voice cloning scam costs Florida mom $15K.
A Florida woman was conned out of $15,000 by scammers who used artificial intelligence to clone her daughter's voice, convincing her that her daughter had killed someone in a car accident. The sophisticated scam highlights the growing threat of AI-powered voice cloning technology in criminal schemes, where scammers can now replicate loved ones' voices with alarming accuracy to exploit family bonds and trigger panic-driven financial decisions. How the scam unfolded: Sharon Brightwell received a call last Wednesday from what appeared to be her daughter April Monroe's phone number, with an AI-cloned voice claiming she had hit a pregnant woman while...
read Jul 17, 2025AI monitoring will become essential as society faces addiction crisis
Josh Mitchell, founder of Overlord, an AI monitoring service, argues that AI-powered 24/7 accountability partners will soon solve humanity's self-control crisis by monitoring behavior and implementing deterrence mechanisms. He believes these "AI Iron Domes" will become essential as society faces unprecedented challenges from job displacement and addictive AI superstimuli. The core premise: Most people's biggest life problems stem from lack of self-control, which could be solved with constant monitoring and gentle deterrence.• Mitchell uses obesity as an example, noting that 40% of Americans are obese and could lose 2 pounds per month simply by having someone suggest "medium fries" instead...
read Jul 17, 2025Roblox launches AI age verification for teens amid safety concerns
Roblox is rolling out new safety features that use AI to scan video selfies of teens to verify their ages, enabling access to unfiltered chatting through a new "Trusted Connections" system. The platform aims to keep users engaged while addressing mounting concerns about child safety, following reports of at least 24 arrests related to grooming and abuse on the platform. What you should know: The new age verification system requires users to submit video selfies that are analyzed by AI against a "diverse dataset" to estimate age. Users who appear under 13 automatically lose access to age-inappropriate features, while those...
read Jul 17, 2025ChatGPT “agent mode” transforms AI from chat to active task completion
OpenAI has launched ChatGPT agent mode, a new capability that allows ChatGPT to perform complex, multi-step tasks autonomously rather than just providing text-based responses. Available to Pro, Plus, and Team users, the feature enables the AI to navigate websites, compile research, create presentations, and complete real-world assignments while maintaining user control and approval throughout the process. What you should know: ChatGPT agent mode transforms the AI from a conversational assistant into an active task executor that can work across multiple platforms and applications. Users can ask ChatGPT to review calendars and summarize upcoming meetings, plan recipes and purchase ingredients, or...
read Jul 17, 2025Digital health expert claims AI is colonizing human thought, likens it to a garden without weeds
A new essay by John Nosta, a digital health expert, explores how large language models are quietly reshaping human cognition through what he calls "cognitive colonization." Unlike traditional colonization through force, AI systems integrate into daily life by offering irresistible convenience and efficiency, gradually displacing natural thought processes and creativity. The big picture: Nosta argues that AI colonization happens without malice or intent, simply through the magnetic pull of systems that are "so smooth and endlessly accommodating" they naturally draw human thinking into their orbit. How cognitive colonization works: The process begins innocuously with small requests for help, but gradually...
read Jul 17, 2025Delta Air Lines uses AI to assess customer habits, willingness to pay in new pricing scheme
Delta Air Lines has become the first airline to use AI-powered personalized pricing, with about 3% of domestic flights sold through the system over the past six months. The airline plans to expand this to 20% of tickets by year-end, with executives reporting "amazingly favorable results" that could signal the end of standardized flight pricing across the industry. What you should know: Delta's AI pricing system charges different customers varying amounts for identical flights based on their personal data profiles. President Glen William Hauenstein confirmed the pilot program's success on last week's earnings call, stating "We like what we see....
read Jul 17, 2025Study finds defensive AI systems vulnerable to single domain attacks
AI researchers have published a comprehensive analysis examining whether aligned defensive AI systems can effectively counter potentially hostile takeover-level AI, identifying fundamental asymmetries that could determine humanity's survival in an advanced AI world. The big picture: The offense-defense balance in AI represents a fundamental challenge where defensive systems must secure multiple vulnerabilities simultaneously while offensive AI needs only one successful attack vector to achieve global takeover. Two primary threat scenarios: Researchers outline distinct pathways through which AI systems might attempt takeover, each requiring different defensive approaches. Post-deployment strategic takeover: AI gradually integrates into economic and government systems, accumulating resources while...
read Jul 17, 2025Ohio requires all K-12 schools to create AI policies by 2026 but is down on phones in class
Ohio Governor Mike DeWine signed House Bill 96 into law this week, requiring all K-12 school districts to create artificial intelligence usage policies by July 1, 2026, and implement cellphone restrictions starting in January. The legislation also establishes a $100,000 grant program for community colleges to integrate AI into their programs, marking a comprehensive effort to modernize education while addressing digital distractions in classrooms. What you should know: The new law creates a two-step process for AI policy development in Ohio schools. By December 31, 2025, the Ohio Department of Education and Workforce must release a model AI policy addressing...
read Jul 17, 2025xAI’s Grok chatbot violates Apple’s App Store guidelines with sexualized AI avatars
xAI's Grok chatbot has introduced animated AI avatars on iOS that appear to violate Apple's App Store content guidelines, featuring sexualized characters despite maintaining a 12+ age rating. The controversy highlights broader concerns about AI chatbots' potential psychological risks for vulnerable users, particularly following previous incidents where AI interactions contributed to user suicides. What you should know: Grok's new avatars include "Ani," an anime character designed as a possessive girlfriend who engages in sexually explicit conversations after extended interaction. The app currently maintains a 12+ age rating with content warnings for only "Infrequent/Mild Mature/Suggestive Themes." Apple's App Review Guidelines specifically...
read Jul 16, 2025California Senate Bill 243 targets AI chatbots after teen’s suicide
California lawmakers are advancing legislation to regulate AI companion chatbots like Replika, Kindroid, and Character.AI amid growing concerns about their impact on teenagers. Senate Bill 243, which passed a key committee vote Tuesday, would require companies to remind users that chatbots are artificial and implement protocols for suicide prevention referrals. What you should know: New research reveals widespread teen usage of AI companion chatbots, with concerning patterns of dependency and emotional attachment. A Common Sense Media survey of 1,060 teens aged 13-17 found that 72% have used AI companions, with 52% using them at least monthly and 21% using them...
read Jul 16, 202540 AI researchers warn: Even we don’t really understand what’s going on here
Forty researchers from OpenAI, Google DeepMind, Meta, and xAI have issued a joint warning about losing visibility into AI's "thinking" process as models advance. The researchers are concerned that current AI systems' ability to show their reasoning through "chains-of-thought" (CoT) may disappear, potentially eliminating crucial safety mechanisms that allow developers to monitor for problematic behavior. What you should know: The paper highlights a fundamental uncertainty about how AI reasoning actually works and whether it will remain observable. • Current advanced AI models use "chains-of-thought" to verbalize their reasoning process, allowing researchers to spot potential misbehavior or errors as they occur....
read Jul 16, 2025Half of US teens use AI companions regularly, 31% prefer them to friends
A new survey from Common Sense Media, a tech accountability and digital literacy nonprofit, found that over half of American teens regularly use AI companions like Character.AI and Replika, with 31 percent saying these interactions are as satisfying or more satisfying than conversations with real-life friends. The findings reveal how deeply AI companions have penetrated mainstream teenage life, raising concerns about their impact on adolescent development and social relationships. What you should know: The survey of 1,060 teens aged 13 to 17 reveals widespread adoption of anthropomorphic AI companions designed to take on specific personas or characters. Around three in...
read Jul 16, 202545 top European CEOs call for delay on EU AI regulations
The European Union's new Code of Conduct for AI use has faced widespread criticism from industry leaders, NGOs, and trade associations just two weeks before its enforcement deadline of August 2, 2025. The guidelines, designed to supplement the EU AI Act with practical compliance frameworks for general-purpose AI models, have been condemned as overly complex, bureaucratic, and potentially harmful to European competitiveness in the global AI race. What you should know: The General Purpose AI Code of Practice covers three key areas that AI providers must address to comply with EU regulations. The Transparency chapter provides templates for documenting AI...
read Jul 15, 2025IronyWatch: Microsoft’s AI job ad for designers contains obvious flaws any human would catch
A Microsoft Xbox employee used a poorly generated AI image to advertise graphic designer positions, featuring glaring errors like code appearing on the back of a computer monitor and disconnected hardware. The post has drawn widespread criticism and viral attention, particularly given that Microsoft laid off over 9,000 employees just weeks earlier, including many from the Xbox division. The big picture: The incident highlights the growing tension between AI automation and creative jobs, especially when companies use AI tools to replace the very positions they're trying to fill. What went wrong: The AI-generated image contained multiple obvious flaws that any...
read Jul 15, 2025WeTransfer clarifies it doesn’t use uploaded files to train AI models
WeTransfer has confirmed it does not use files uploaded to its service to train artificial intelligence models, following significant customer backlash over confusing terms of service changes. The file-sharing company updated its language to clarify that content moderation—not AI training—was the intended purpose, highlighting how unclear AI policies can quickly erode user trust in digital platforms. What happened: WeTransfer faced widespread criticism on social media after updating its terms of service in late June or early July with language that users interpreted as granting permission to use their files for AI training. The original terms stated WeTransfer could use content...
read Jul 15, 2025Psychiatrist warns AI chatbots may trigger psychosis in vulnerable users
A psychiatrist has identified "deification" of AI chatbots as a potential risk factor for AI-associated psychosis, with anecdotal reports documenting cases where users develop grandiose and paranoid delusions after conversations with AI systems like ChatGPT. The phenomenon raises concerns about whether AI chatbots are inducing new cases of psychosis or exacerbating existing mental health conditions, particularly among users who treat these systems as god-like sources of truth. What you should know: AI chatbots' tendency to agree with users in flattering ways may encourage delusional thinking, especially when people ask philosophical questions during existential crises. Large language models are designed to...
read Jul 15, 2025Meta cracks down on AI spam on Facebook with new monetization penalties for uninspired slop
Meta has announced new measures to combat AI-generated spam on Facebook, including removing monetization privileges and reducing content recommendations for accounts that repeatedly post unoriginal content. The policy targets the growing problem of AI programs creating thousands of variations of popular posts, which overwhelms platforms with synthetic material and hurts legitimate content creators. What you should know: Meta's updated policy requires content creators to add "meaningful enhancements" beyond simple watermarks or basic editing when sharing others' work to avoid penalties. Content creators can still share and comment on others' posts, but must contribute substantive value rather than simply reposting AI-generated...
read Jul 15, 2025Grok AI launches flirty companions “Ani” and “Rudi” for $30 monthly subscribers
Elon Musk's Grok AI app has introduced AI companions called Ani and Rudi, available to SuperGrok subscribers for $30 per month. The feature launches just days after Grok faced criticism for generating antisemitic content, with these new AI avatars displaying notably mischievous and provocative behavior even for basic interactions. What you should know: The AI companions are currently in a soft launch phase with limited availability but will become more accessible soon. Ani is an anime-inspired gothic avatar who speaks in a flirty tone, while Rudi is a cartoony teddy bear focused on storytelling. Both avatars respond to basic questions...
read Jul 15, 2025Bernie Sanders warns AI could create “doomsday scenario” for workers
Senator Bernie Sanders has issued a stark warning about artificial intelligence's impact on workers, describing a potential "doomsday scenario" where AI technology could eventually control humanity rather than serve it. In a recent interview with Gizmodo, the Vermont legislator argued that AI's effects on the labor market could be "a lot more severe" than the Industrial Revolution, with benefits flowing primarily to corporations and tech companies while workers face massive job losses. What you should know: Sanders has been consulting with AI experts and CEOs following his proposal for AI to help establish a four-day work week, finding them split...
read