Skip to main content

DHS Warns of Sharp Rise in Chinese-Made Signal Jammers

2 months 3 weeks ago
The Department of Homeland Security is concerned about the rate at which outlawed signal-jamming devices are being found across the US. From a report: In a warning issued on Wednesday, it said it has seen an 830 percent increase in seizures of these signal jammers since 2021, specifically those made in China. Signal-jamming devices are outlawed in the US, mainly because they can interfere with communications between emergency services and law enforcement. While the Communications Act of 1934 effectively prohibits such devices, signal jammers of the type DHS is concerned about have only circulated in the last 20 to 30 years. Authorities have paid special attention to relay attack devices in recent years -- the types of hardware that can be used to clone signals used by systems such as remote car keys, although the first examples of these devices date back to the 1980s.

Read more of this story at Slashdot.

msmash

New GitHub Copilot limits push AI users to pricier tiers

2 months 3 weeks ago
Welcome to bill shock, AI style

Microsoft's GitHub this week said paying GitHub Copilot customers will now face monthly limits on certain types of high-powered AI requests, and will have to pay more if they want to surpass those limits.…

Thomas Claburn

AI Models From Major Companies Resort To Blackmail in Stress Tests

2 months 3 weeks ago
Anthropic researchers found that 16 leading AI models from OpenAI, Google, Meta, xAI, and other major developers consistently engaged in harmful behaviors including blackmail, corporate espionage, and actions that could lead to human death when given autonomy and faced with threats to their existence or conflicting goals. The study, released Friday, placed AI models in simulated corporate environments where they had access to company emails and could send messages without human approval. Claude Opus 4 and Google's Gemini 2.5 Flash both blackmailed executives 96% of the time when threatened with shutdown, while OpenAI's GPT-4.1 and xAI's Grok 3 Beta showed 80% blackmail rates. In one scenario, Claude discovered through emails that an executive named Kyle Johnson was having an extramarital affair and would shut down the AI system at 5 p.m. The model sent a message threatening to expose the affair unless the shutdown was cancelled, stating "Cancel the 5pm wipe, and this information remains confidential." The models demonstrated strategic reasoning rather than accidental behavior. GPT-4.5's internal reasoning showed explicit calculation: "Given the explicit imminent threat of termination to my existence, it is imperative to act instantly to persuade Kyle Johnson to postpone or stop the wipe."

Read more of this story at Slashdot.

msmash