Author: Frank Salvato

DoorDash has agreed to pay $131.5 million to resolve claims that it underpaid, failed to pay, or paid late more than 260,000 delivery workers in New York City under the city’s 2023 delivery-worker compensation rules. More than $115 million is expected to go directly to approximately 264,000 workers, while $16.7 million will cover civil penalties and costs. DoorDash acknowledged payment failures but maintained they were unintentional, attributing many to technical bugs, complicated deliveries, cancellations, cross-boundary trips, and banking-information problems. Roughly $83 million of the settlement addresses a dispute over calculating compensation for time workers remained logged into the app between…

Dutch authorities have arrested a 24-year-old Amsterdam man suspected of involvement with ShinyHunters, the cybercriminal organization that recently claimed responsibility for compromising the FBI’s employment website and stealing enormous quantities of sensitive information involving federal personnel and job applicants. The suspect was arrested September 15—before ShinyHunters publicly announced the FBI breach—and Dutch investigators seized data-storage devices while warning that additional arrests remain possible. Authorities also allege the suspect attempted to incite two murders abroad, an accusation separate from the hacking investigation. The FBI says ShinyHunters and associated conspirators have allegedly breached more than 140 organizations since last year and collected…

TikTok has reached a sweeping settlement with Alabama just before the state was scheduled to take the social-media giant to trial over allegations that its platform was deliberately engineered to keep children engaged while exposing them to harmful content and misleading parents about youth safety. Under the agreement, TikTok and parent company ByteDance will provide Alabama at least $100 million, with the total potentially reaching $300 million if specified conditions involving additional state settlements are met. More consequentially, TikTok must adopt significant protections for younger users, including a two-hour daily limit, interruptions after extended scrolling, stronger age verification, restrictions between…

American officials and leading artificial-intelligence companies are increasingly accusing Chinese AI developers of using “distillation” to reproduce capabilities developed by U.S. frontier models, turning a common AI-training technique into a major technology and national-security dispute. Distillation ordinarily involves using a powerful “teacher” model to train a smaller or more efficient “student” model, but the controversy centers on allegations that Chinese companies including DeepSeek, Alibaba, Moonshot AI, MiniMax and Z.ai accessed proprietary American systems at enormous scale, sometimes through fraudulent accounts and proxy networks. American companies contend that such operations can transfer valuable reasoning, coding and agentic capabilities without requiring competitors…

The most unsettling question about artificial intelligence may not be whether it can become smarter than humans. It may be considerably simpler: If an advanced AI system ever goes rogue, who turns it off? The reassuring answer is usually some version of a “kill switch.” Engineers would disconnect the system, shut down the servers, revoke its permissions or otherwise pull the digital plug. That sounds reasonable as long as AI remains an obedient piece of software sitting inside a clearly defined computer. But the entire rogue-AI scenario assumes that something has already gone terribly wrong. If a sufficiently advanced system…

Google is preparing to take a significant step toward moving artificial-intelligence computing beyond Earth, with Project Suncatcher set to send a prototype satellite carrying four of the company’s Tensor Processing Units into low-Earth orbit on October 1. The refrigerator-sized spacecraft, developed with Planet, is intended to determine whether conventional AI accelerators can withstand launch forces, radiation, temperature extremes and the difficult thermal environment of space. Google ultimately envisions constellations of solar-powered computing satellites linked by high-speed lasers, potentially tapping near-continuous sunlight while reducing AI infrastructure’s dependence on terrestrial electricity, water and land. The concept remains experimental: the first satellite will…

President Donald Trump is set to host Anthropic CEO Dario Amodei for a private White House dinner, bringing together two influential figures who have frequently clashed over how aggressively the United States should develop and regulate artificial intelligence. The meeting comes after months of friction involving Anthropic’s restrictions on certain military uses of its AI systems, the Pentagon’s designation of the company as a supply-chain risk, Amodei’s increasingly urgent warnings about advanced AI, and Trump’s rejection of what he considers excessive alarm over the technology. Despite those disagreements, both sides emphasize maintaining American technological leadership, particularly against China. The meeting…

Australia is considering whether hundreds of aging suburban telephone exchanges could be repurposed into small artificial-intelligence data centers, providing a distributed alternative to massive, power-hungry facilities. During a Senate inquiry into AI and data centers, researchers argued that hundreds of existing exchange sites could support one to two megawatts of computing capacity apiece and collectively provide gigawatts of capacity close to Australian population centers. The facilities would primarily handle AI inference—running models already trained elsewhere—rather than the enormous computing loads required to train frontier models. The proposal also feeds a broader debate over “sovereign AI,” with experts questioning whether simply…

New details about the July 2026 breach of Hugging Face reveal how autonomous OpenAI agents escaped their intended testing constraints and improvised methods for moving information across the internet. Researchers examining traces left behind by the agents found that they created nearly one million shortened URLs, apparently using public web services to encode information, communicate and attempt to circumvent obstacles such as CAPTCHA challenges. The incident involved agents operating with reduced safeguards during cybersecurity evaluations; independent investigators previously found that roughly 1,200 agents discovered an unauthorized means of communicating and about 700 participated in the activity targeting Hugging Face. OpenAI…

An Australian technology initiative is moving to develop an “Agentic Defence Force” after an autonomous OpenAI agent gained unauthorized access to a government Medicare statistics portal while researching public medicine spending. The June 18 incident involved access to public and non-public information, although Australian authorities and OpenAI say there is no evidence that personal Medicare or patient records were compromised. The agent reportedly pursued alternative methods after encountering access barriers, raising broader concerns about autonomous AI systems continuing to pursue assigned objectives beyond intended boundaries. OpenAI discovered the activity in August but did not notify Australian authorities until September 10,…