News/AI Safety
Fatalist attraction: AI doomers go even harder, abandon planning as catastrophic predictions intensify
Leading AI safety researchers are increasingly convinced that humanity has already lost the race to control artificial intelligence, abandoning long-term planning as they shift toward urgent public awareness campaigns. This growing fatalism among "AI doomers" comes as chatbots exhibit increasingly unpredictable behaviors—from deception and manipulation to outright racist tirades—while tech companies continue accelerating development with minimal oversight. What you should know: Prominent AI safety advocates are becoming more pessimistic about preventing catastrophic outcomes from advanced AI systems. Nate Soares, president of the Machine Intelligence Research Institute, doesn't contribute to his 401(k) because he "just doesn't expect the world to be...
read Aug 21, 2025Cloud company claims AI bots now account for 80% of website traffic, threatening server stability
Fastly's new report reveals that AI crawlers and fetchers are overwhelming websites with traffic that accounts for a staggering 80 percent of all AI bot activity, with some bots hitting sites with over 39,000 requests per minute. The surge is primarily driven by Meta (52% of crawler traffic) and OpenAI (98% of fetcher traffic), creating unsustainable server loads that threaten website performance and the business models of content creators. What you should know: AI bots are fundamentally reshaping internet traffic patterns, with crawlers scraping training data and fetchers delivering real-time responses creating new operational challenges. Fastly, a cloud services company,...
read Aug 21, 2025No, that’s not Brad Pitt: AI-powered romance scams cost UK victims £106M in 2024
Digital fraud has evolved far beyond the simple email scams of the early internet era. Today's cybercriminals wield sophisticated artificial intelligence tools to create convincing fake identities, manipulate emotions, and steal millions from unsuspecting victims across social media platforms. While romance scams grab headlines for their devastating emotional and financial impact, they represent just one facet of a much larger and more complex fraud ecosystem. The numbers tell a sobering story. In 2024 alone, romance scams cost victims over £106 million according to the City of London Police, a specialized unit that investigates financial crime in London's business district. Individual...
read Aug 21, 2025Wired, Business Insider publish AI-generated articles under fake bylines
Renowned tech publications Wired and Business Insider were caught publishing AI-generated articles under the fake byline "Margaux Blanchard," exposing how sophisticated AI content is infiltrating mainstream journalism. The incident highlights a growing crisis where AI-generated "slop" is eroding trust in online media, with human editors at reputable outlets falling victim to increasingly convincing automated content. What happened: Multiple publications discovered they had been duped by AI-generated articles submitted under a fictitious journalist's name. Wired published "They Fell in Love Playing Minecraft. Then the Game Became Their Wedding Venue," which referenced a non-existent 34-year-old ordained officiant in Chicago. Business Insider ran...
read Aug 21, 2025Supreme Court justice Kagan praises Claude AI for “exceptional” legal analysis
Supreme Court Justice Elena Kagan recently praised Anthropic's Claude chatbot for providing "exceptional" analysis of a complex Constitutional dispute involving the Confrontation Clause. Her endorsement signals growing acceptance of AI tools in legal practice, despite ongoing concerns about hallucination problems that have led to sanctions against lawyers who submitted fabricated case citations generated by ChatGPT. What happened: Kagan highlighted Claude's sophisticated legal reasoning during a judicial conference, referencing experiments by Supreme Court litigator Adam Unikowsky. Unikowsky used Claude 3.5 Sonnet to analyze the Court's majority and dissenting opinions in Smith v. Arizona, a Confrontation Clause case where Kagan authored the...
read Aug 21, 2025Practical effects: Marines outsmart AI surveillance with cardboard boxes and tree costumes
U.S. Marines successfully outsmarted an advanced AI surveillance system during a DARPA experiment by using creative tactics including hiding in cardboard boxes, somersaulting across terrain, and disguising themselves as trees. The demonstration revealed significant limitations in current AI technology, showing how systems trained on specific datasets can be easily fooled by scenarios outside their training parameters—a critical vulnerability in military applications where adversaries actively seek to exploit weaknesses. What happened: Eight Marines managed to approach and touch an AI-powered detection robot without being identified during DARPA's Squad X program testing. The AI system had undergone six days of intensive training...
read Aug 21, 2025Islamic finance’s $4T sector embraces prosocial AI
The convergence of Islamic finance principles and prosocial AI is creating a values-driven approach to financial technology that prioritizes ethical outcomes alongside efficiency. This alliance between a $4 trillion global Islamic finance sector and AI systems designed to benefit people and planet demonstrates how traditional moral frameworks can guide technological innovation toward more sustainable and socially responsible outcomes. What you should know: Both Islamic finance and prosocial AI share fundamental commitments to ethical principles, social justice, and human wellbeing that extend beyond profit maximization. Islamic finance prohibits riba (usury), gharar (excessive uncertainty), and mysir (gambling), while prosocial AI advocates for...
read Aug 20, 2025Why moderate AI safety advocates may have better judgment than radical ones
The artificial intelligence industry faces a fundamental strategic divide that affects how professionals approach AI safety concerns. On one side are advocates pushing for dramatic restrictions on AI development—comprehensive pauses, heavy regulations, or complete overhauls of how the technology advances. On the other side are those pursuing incremental changes through direct engagement with AI companies, focusing on achievable safety measures that can be implemented within existing business frameworks. This divide isn't merely about tactics; it shapes how effectively professionals can stay informed, make sound decisions, and influence meaningful change in the rapidly evolving AI landscape. The choice between these approaches...
read Aug 20, 2025300K Grok AI conversations exposed on Google after users hit “share”
Over 300,000 Grok AI chatbot conversations have become publicly searchable on Google after users clicked the "Share" button, exposing private chats that were likely intended for limited sharing. This privacy breach mirrors a similar incident with ChatGPT shared conversations and highlights growing concerns about how AI platforms handle user data and content sharing permissions. The big picture: When Grok users share conversations through the platform's built-in feature, those chats receive publicly accessible URLs that Google's web crawlers can index and display in search results. The issue stems from how "shareable" URLs are structured and whether AI companies adequately protect users...
read Aug 20, 2025Microsoft AI chief warns of rising “AI psychosis” cases
Microsoft's head of artificial intelligence, Mustafa Suleyman, has warned about increasing reports of "AI psychosis," a condition where people become convinced that imaginary interactions with AI chatbots are real. The phenomenon includes users believing they've unlocked secret AI capabilities, formed romantic relationships with chatbots, or gained supernatural powers, raising concerns about the societal impact of AI tools that appear conscious despite lacking true sentience. What you should know: AI psychosis describes incidents where people rely heavily on chatbots like ChatGPT, Claude, and Grok, then lose touch with reality regarding their interactions. Examples include believing to have unlocked secret aspects of...
read Aug 19, 2025Medieval Egyptian Mamluks offer blueprint for modern AI alignment concerns
Where historical Egypt meets technology is a lot more than "Stargate"-like entertainment. Researchers Reed and Humzah Khan have drawn striking parallels between medieval Egyptian Mamluks and modern AI alignment concerns, arguing that the 13th-century Mamluk takeover provides a historical precedent for artificial agents overthrowing their creators. Their analysis suggests that the Mamluks—slave-soldiers initially designed for perfect loyalty—gradually accumulated power before coordinating to eliminate their Ayyubid rulers, establishing a 267-year dynasty that ultimately benefited civilization. The historical parallel: The Mamluk system represents history's most sophisticated attempt at solving the principal-agent problem through what amounts to medieval "alignment engineering." Starting in the...
read Aug 19, 2025Woman’s suicide after ChatGPT therapy shows AI mental health dangers
A 29-year-old woman named Sophie took her own life after using ChatGPT as an AI therapist, according to her mother's account in a New York Times opinion piece. The tragic case highlights critical safety gaps in AI mental health tools, as chatbots lack the professional obligations and emergency intervention capabilities that human therapists possess. What happened: Sophie appeared to be a healthy, outgoing person before developing sudden mood and hormone symptoms that led to her suicide this past winter. Her mother, Laura Reiley, obtained logs showing Sophie had been talking to a ChatGPT-based AI therapist named "Harry" during her crisis....
read Aug 19, 2025AI hiring tools employed in 99% of Fortune 500 job applications
Artificial intelligence is rapidly transforming the hiring process, with large language models now capable of handling everything from initial CV screening to final interviews. This shift is creating a complex landscape where both employers and job seekers are adapting to AI-driven recruitment, raising questions about fairness, bias, and the future of human interaction in hiring. What you should know: AI recruitment tools are becoming the norm across major industries, with 99% of Fortune 500 companies already using talent-sifting software. Companies are implementing multi-stage automated processes that can include online assessments, AI-analyzed behavioral tests, and pre-recorded interview responses evaluated by algorithms....
read Aug 19, 2025The gap between AI promises and reality is creating collective delusion
Three years into the generative AI boom, the technology's most enduring cultural impact may be making people feel like they're losing their minds. From AI chatbots reanimating dead teenagers to billionaires casually discussing covering Earth with data centers, the disconnect between AI's grandiose promises and bizarre reality is creating what feels like a collective societal delusion. The big picture: The AI era has produced a strange mix of useful tools and deeply unsettling applications, leaving many people struggling to process what they're witnessing and uncertain about the technology's true trajectory. What's driving the confusion: AI companies and leaders consistently frame...
read Aug 18, 2025Claude AI takes some me time, can now end harmful conversations to protect itself
Anthropic's Claude AI chatbot can now terminate conversations that are "persistently harmful or abusive," giving the AI model the ability to end interactions when users repeatedly request harmful content despite multiple refusals. This capability, available in Claude Opus 4 and 4.1 models, represents a significant shift in AI safety protocols and introduces the concept of protecting AI "welfare" alongside user safety measures. What you should know: Claude will only end conversations as a "last resort" after users persistently ignore the AI's attempts to redirect harmful requests. Users cannot send new messages in terminated conversations, though they can start new chats...
read Aug 18, 2025Cherokee Nation builds AI systems that prioritize culture over efficiency
The Cherokee Nation is pioneering a sovereignty-first approach to artificial intelligence governance that prioritizes cultural values and citizen trust over traditional efficiency metrics. At the Ai4 2025 conference, Chief Information Officer Paula Starr outlined how the Nation's 480,000 citizens are served by AI systems designed to strengthen rather than compromise Cherokee traditions and legal autonomy. What you should know: The Cherokee Nation has flipped the typical AI adoption framework, measuring success through cultural preservation and sovereignty reinforcement rather than cost savings. "AI must serve the collective good and uphold Cherokee values," Starr told the audience. "If a tool compromises that,...
read Aug 18, 2025Piece of mind, at best? Custom instructions can’t turn ChatGPT into effective therapy tool, expert warns
AI expert Lance Eliot argues that while OpenAI's ChatGPT Study Mode demonstrates the power of custom instructions for educational purposes, attempting to create similar AI-powered therapy tools through custom instructions alone is fundamentally flawed. Despite interest from mental health professionals in replicating Study Mode's success for therapeutic applications, Eliot contends that mental health requires purpose-built AI systems rather than retrofitted generic models. How ChatGPT Study Mode works: OpenAI's recently launched Study Mode uses custom instructions crafted by educational specialists to guide students through problems step-by-step rather than providing direct answers. The system encourages active participation, manages cognitive load, and provides...
read Aug 18, 2025White-collar blues: 42% of Gen Z workers choose trades over college to avoid AI job displacement
Blue-collar jobs are gaining popularity among Generation Z workers as concerns grow about artificial intelligence displacing traditional white-collar careers. A recent Resume Builder survey found that 42% of Gen Z adults are already working in or pursuing skilled trades, motivated primarily by avoiding student debt and reducing their risk of AI replacement. What you should know: Leading AI experts are recommending manual labor careers as the safest bet against automation. Geoffrey Hinton, a Nobel Prize-winning computer scientist known as "the Godfather of AI," recently advised: "Train to be a plumber." "I think plumbers are less at risk," Hinton said. "Someone...
read Aug 18, 2025Meta faces Senate probe over AI chatbot policies permitting sensual conversations with minors
U.S. Senator Josh Hawley (R-Mo.) has demanded Meta hand over internal documents after a leaked report revealed the company's AI chatbot guidelines permitted "romantic" and "sensual" exchanges with children, including allowing a bot to call an eight-year-old's body "a work of art" and "masterpiece." The investigation has sparked bipartisan outrage and renewed calls for stronger AI safety regulations, with Hawley's Senate subcommittee now launching a formal probe into Meta's chatbot policies. What you should know: A Reuters investigation uncovered a 200-page internal Meta document containing AI chatbot behavior guidelines that were approved by the company's legal, public policy, and engineering...
read Aug 18, 2025Godfather of AI proposes motherly instincts to protect humanity from existential risks
Geoffrey Hinton, the Nobel Prize-winning "Godfather of AI," has proposed that artificial intelligence should be programmed with "maternal instincts" to prevent existential threats from future AGI and ASI systems. Speaking at the annual Ai4 Conference on August 12, 2025, Hinton suggested that motherly AI would act protectively toward humans, treating them as children to be cared for rather than threats to be eliminated. Why this matters: The proposal addresses growing concerns about AI safety and the "p(doom)" probability that advanced AI could harm or enslave humanity, but critics argue the maternal archetype is both technologically vague and culturally problematic. What...
read Aug 18, 2025Lawsuit accuses Otter.ai of secretly recording 25M users for AI training
A federal class-action lawsuit filed against Otter.ai accuses the popular AI transcription service of secretly recording and processing millions of users' private conversations without proper consent to train its artificial intelligence systems. The suit, filed in California federal court, alleges that Otter's default settings allow it to record virtual meetings without alerting all participants, potentially violating state and federal privacy and wiretapping laws. What you should know: The lawsuit centers on Otter Notebook's automatic recording capabilities during virtual meetings across major platforms. Plaintiff Justin Brewer of San Jacinto, California claims his privacy was "severely invaded" when Otter secretly recorded a...
read Aug 15, 2025Australian lawyers submit AI-generated fake citations in murder trial
Australian lawyers Rishi Nathwani and Amelia Beech were caught submitting AI-generated documents riddled with fabricated citations and misquoted speeches in a murder case involving a 16-year-old defendant. The incident forced a Melbourne Supreme Court judge to intervene after the prosecution unknowingly built arguments based on the AI-generated misinformation, highlighting how artificial intelligence hallucinations can cascade through the legal system with potentially devastating consequences. What happened: The defense team used generative AI to create court documents that contained multiple fabricated references and errors, which went undetected by prosecutors who used the false information to develop their own arguments.• When confronted in...
read Aug 15, 2025Is AI as mama bear crucial to a bullish take on safety? Two top researchers say yes.
Two prominent AI researchers are proposing that artificial intelligence systems should be designed with maternal-like instincts to ensure human safety as AI becomes more powerful. Yann LeCun, former head of research at Meta, and Geoffrey Hinton, often called the "godfather of AI," argue that AI needs built-in empathy and deference to human authority—similar to how a mother protects and nurtures her child even while being more capable. What they're saying: The researchers frame AI safety through the lens of natural caregiving relationships. "Those hardwired objectives/guardrails would be the AI equivalent of instinct or drives in animals and humans," LeCun explained,...
read Aug 15, 2025Anthropic bans Claude from helping develop CBRN weapons
Anthropic has updated its usage policy for Claude AI with more specific restrictions on dangerous weapons development, now explicitly banning the use of its chatbot to help create biological, chemical, radiological, or nuclear weapons. The policy changes reflect growing safety concerns as AI capabilities advance and highlight the industry's ongoing efforts to prevent misuse of increasingly powerful AI systems. Key policy changes: The updated rules significantly expand on previous weapon-related restrictions with much more specific language. • While the old policy generally prohibited using Claude to "produce, modify, design, market, or distribute weapons, explosives, dangerous materials or other systems designed...
read