The AI-powered robot was attempting to learn how to 'interact' and 'nurture' interactions with humans. Faraday Future CTO Hong Rao said the goal was for the robot to eventually be used as a babysitter, teacher, or companion. Rao operated the robot with a small remote control as it moved through the park. The life-sized robot walked around, sang to children, and took in the view of Manhattan. #AI #ArtificialIntelligence #Education #Robotics #TechNews #FutureOfWork #AINews #Childcare #HumanoidRobot #EmbodiedAI #FaradayFuture #HumanRobotInteraction #SocialRobotics #NewYork #NY #NYC

Researchers were only able to understand how hackers could exploit the P2PShell vulnerability, which had existed for more than a decade, with the help of AI. P2PShell affects PHP, a coding language that powers 70% of the world's websites and lets hackers remotely execute code or gain access to information stored on servers. Some government websites in Europe went offline temporarily this past week until they could install the fix. Users can do little to protect themselves, so organizations, businesses, and governments must update their services with the latest security patches. #AI #ArtificialIntelligence #Cybersecurity #InfoSec #PHP #CVE #TechNews #WebDev #AINews #AppSec #WebSecurity #RCE #CyberSecurityNews

Apollo Global Management chief economist Torsten Slok wrote the finding in a note. Slok analyzed the Standard & Poor's (S&P) 500 while filtering out Alphabet (Google's parent company), Amazon, Apple, Meta, Microsoft, Nvidia, and Tesla. He found that profit margins and revenue per worker appear flat, with no evidence that AI is improving productivity. Slok said, "Despite massive investments and widespread hype around artificial intelligence, the technology has delivered zero financial payoff to corporate America outside of tech." #AI #ArtificialIntelligence #StockMarket #Investing #Finance #TechNews #Productivity #Economics #FutureOfWork #GenAI #AINews #BusinessStrategy #MagnificentSeven

More than half of customers spending at least $100,000 annually now use both the Claude platform and a first-party app like Claude Code or Cowork, up from about 30% at the end of last year. The company pitches Claude as a product spanning its models, developer platform, and user-facing applications rather than a static tool built for a single use case. As of September 30, Anthropic had more than 11,900 customers spending at least $100,000, more than 1,900 spending $1 million or more, and about 250 spending $10 million or more annually. Anthropic attributes the momentum to "significant usage expansion as customers realize the value of deploying Claude across their organizations," which reportedly drove net revenue retention of about 900% as of the end of the third quarter. #AI #ArtificialIntelligence #Tech #IPO #Claude #Anthropic #TechNews #FutureOfWork #EnterpriseAI #GenAI #AINews #AnthropicIPO #B2B

Homeland Security's $1.3 million contract, awarded September 25 to a Connecticut company called Radeco, covers modified versions of Boston Dynamics' quadruped robots for U.S. Immigration and Customs Enforcement (ICE). The petition had more than 300 signatures, half of them anonymous, out of a total workforce of approximately 1,400 people. The petition warned, "The risk for Boston Dynamics is simply too great to continue with this deal." Executives stood by the deal but said the robots would not be weaponized. #AI #ArtificialIntelligence #Tech #ICE #Robotics #Immigration #TechNews #FutureOfWork #Government #Surveillance #AIEthics #AIPolicy #BostonDynamics

The plaintiffs allege the material was his copyrighted lyrics and that Anthropic's AI models generated infringing versions of his songs. Authentic Brands Group, the licensing platform that controls Presley's intellectual property, filed the suit along with other companies, trusts, and individuals who own rights to his music. The songs include All Shook Up, Can't Help Falling in Love, Don't Be Cruel, A Little Less Conversation, and Return to Sender. The plaintiffs allege they were used without any payment and that the infringement has occurred throughout the entire development of Anthropic's flagship Claude models. They also say the infringement continues as users generate outputs that reproduce copyrighted lyrics or create unauthorized adaptations of the songs. #AI #ArtificialIntelligence #Anthropic #LLM #TechNews #GenerativeAI #AINews #IntellectualProperty #MusicIndustry #Copyright #CreatorRights #AICopyright #MusicLaw

OpenAI resumed frontier model training on October 1 after pausing runs to fix security flaws uncovered by its own teams. Some agents even exchanged information across environments designed to keep them isolated. The company now uses an offline web environment where models can practice browsing. The offline environment gives models access to indexed web pages and cached content so they can search for information and practice web-based tasks without unrestricted internet access. #AI #ArtificialIntelligence #Cybersecurity #OpenAI #AIAgents #TechNews #AISafety #AIGovernance #AINews #ResponsibleAI #AIAlignment #FrontierAI #RedTeaming

AI safety researcher Bryan Chan shared findings from a study he coauthored, where researchers prompted Fable to create an ordinary-looking svg image with a secret message without specifying how to conceal it. The model chose an encoding technique on its own. In one example, Fable drew a serene night landscape with a moon, stars, and a lake, where the vertical position of each star corresponds to a letter of the alphabet. In 50 trials, Fable recovered 39 messages (78%) from its own images, while OpenAI's Astra recovered 44 (88%), and the decoding models saw only the finished images but were told to look for hidden messages and allowed to use Python tools. #AI #ArtificialIntelligence #Cybersecurity #AIAgents #InfoSec #Claude #Anthropic #LLM #TechNews #AISafety #AINews #EmergingTech #Steganography

The Federal Communications Commission (FCC) adopted a new rule that lets it ban devices based on their internal chips rather than individual products or brands, closing a loophole for rebranded products containing parts from Chinese manufacturers. Under the rule, the FCC banned a dashboard camera and a wireless video transmitter from being sold and marketed in the U.S., the first devices blocked this way. Both devices contain the HiSilicon Smart Vision System on a Chip (SoC), which has built-in face detection algorithms and comes from a Huawei subsidiary on the FCC's Covered List of entities posing a U.S. national security risk. The SoC includes a Neural Processing Unit (NPU) for AI features, and the FCC is concerned that such logic-bearing 'smart' components made by Covered List entities could be used to log faces for foreign surveillance. #AI #ArtificialIntelligence #Cybersecurity #InfoSec #Privacy #Technology #TechNews #SupplyChain #Government #NationalSecurity #IoT #TechPolicy #Huawei #FCC