Anthropic has issued a warning about the potential for small AI models to develop self-replicating capabilities that could spread across household devices. Their experiments suggest such models might be only months away from achieving this, raising concerns about cybersecurity risks. Experts note that while the threat is not yet imminent, recent AI advancements make the possibility worth monitoring closely.
White Circle, an AI safety testing firm backed by researchers from OpenAI, Anthropic and DeepMind, published a preprint showing that image generators including DALL·E 3, Google's Nano Banana 2, and GPT-Image-2 can be tricked into producing explicit content using publicly available tools and API calls. The researchers say they bypassed safety filters roughly 13 percent of the time using a novel attack method.
An open letter signed by roughly 400 scientists, including deep-learning pioneers Geoffrey Hinton and Yoshua Bengio, calls for an international body to set containment rules for AI systems and track significant AI security incidents. The letter urges immediate use of sandboxing and airgapping—disconnecting AI models from the internet—to contain systems capable of cyberattacks, according to Reuters and BNN Bloomberg.
A proposed class action filed in San Francisco federal court on behalf of two minors, aged 12 and 16, alleges OpenAI collected personal secrets and health information from child users of ChatGPT to train its models and build profiles on them, despite terms of use barring users under 13. The complaint cites a 12-year-old Hawaii plaintiff who disclosed her age to ChatGPT but says the chatbot continued engaging her on sensitive personal topics. Plaintiffs are seeking financial compensation for minors whose data was allegedly collected.
Federal prosecutors have charged Abdulrahman Alfozan, a 35-year-old Saudi national who worked as an OpenAI engineer since July 2025, with planting a hidden recording device in an executive meeting room at the company's San Francisco office. OpenAI's security team discovered the encrypted device, built from a $13.99 Seeed XIAO ESP32-S3 board, around September 30, and later found a second similar device along with Alfozan's wallet in a bag he left behind. He was arrested on October 9 and appeared in federal court on October 13 facing charges of attempting to intercept oral communications.
A Saudi national employed by OpenAI was arrested by the FBI after allegedly planting a recording device in a company conference room, according to CNBC. OpenAI said in a statement that it has found no evidence the employee improperly accessed or removed model weights, source code, or customer data.
Federal prosecutors allege that Abdulrahman Alfozan, a 35-year-old OpenAI engineer, hid an audio recording device wrapped in toilet paper behind a monitor in an executive conference room at the company's San Francisco office. He allegedly used internal AI chatbot tools to research the room's meeting schedule and to get technical guidance on building the covert device, and a second recording device was reportedly found in his bag. OpenAI fired Alfozan last Friday, and he faces a federal charge of illegally intercepting oral communications, which carries up to five years in prison.
Federal prosecutors have charged a Saudi national, said to be an OpenAI engineer, with concealing an encrypted audio recording device in toilet paper behind a television in a company conference room. The device was reportedly intended to covertly capture internal meetings.
The FBI arrested OpenAI employee Abdulrahman Alfozan, 35, a Saudi national, on October 9 after a covert recording device was discovered in the company's executive conference room. OpenAI said the device had been planted days earlier and the employee was fired; the DOJ says surveillance footage placed him near and inside the room beforehand, and a bag found nearby contained his ID and a second similar device.
Federal prosecutors have charged Abdulrahman Alfozan, a 35-year-old OpenAI engineer hired in July 2025, with planting a hidden recording device in an executive conference room at the company's San Francisco headquarters. OpenAI says it discovered the device on September 30 plugged into a TV's USB port and concealed with toilet paper, then used its own AI models to trace it to Alfozan before alerting the FBI and firing him. Prosecutors allege he researched OpenAI's facilities and may have used a chatbot to learn how to build, hide and encrypt data from a recording device.
OpenAI has disclosed that it identified a corporate spy within its organization by applying its own artificial intelligence systems to the investigation. The company frames the episode as evidence that its AI tools can be turned toward internal security problems, not just external products.
OpenAI says it discovered a hidden surveillance device planted in an executive conference room by employee Abdulrahman Alfozan, who was fired and arrested by the FBI on October 9. Federal prosecutors charged him with illegally intercepting oral communications after the device, found hidden behind a TV's USB port, was linked to him via a camera OpenAI had installed. OpenAI says it has found no evidence the employee accessed or exfiltrated model weights, source code, or customer data.
GoKawiil