Linux fréttir

Tells court 'What I did was wrong and I want to apologize for my conduct'

Terraform Labs founder Do Kwon has pled guilty to committing fraud when promoting the so-called "stablecoin" Terra USD and now faces time in jail.…

Cornell University researchers have developed an "invisible" light-based watermarking system that embeds unique codes into the physical light that illuminates the subject during recording, allowing any camera to capture authentication data without special hardware. By comparing these coded light patterns against recorded footage, analysts can spot deepfake manipulations, offering a more resilient verification method than traditional file-based watermarks. TechSpot reports: Programmable light sources such as computer monitors, studio lighting, or certain LED fixtures can be embedded with coded brightness patterns using software alone. Standard non-programmable lamps can be adapted by fitting them with a compact chip -- roughly the size of a postage stamp -- that subtly fluctuates light intensity according to a secret code. The embedded code consists of tiny variations in lighting frequency and brightness that are imperceptible to the naked eye. Michael explained that these fluctuations are designed based on human visual perception research. Each light's unique code effectively produces a low-resolution, time-stamped record of the scene under slightly different lighting conditions. [Abe Davis, an assistant professor] refers to these as code videos. "When someone manipulates a video, the manipulated parts start to contradict what we see in these code videos," Davis said. "And if someone tries to generate fake video with AI, the resulting code videos just look like random variations." By comparing the coded patterns against the suspect footage, analysts can detect missing sequences, inserted objects, or altered scenes. For example, content removed from an interview would appear as visual gaps in the recovered code video, while fabricated elements would often show up as solid black areas. The researchers have demonstrated the use of up to three independent lighting codes within the same scene. This layering increases the complexity of the watermark and raises the difficulty for potential forgers, who would have to replicate multiple synchronized code videos that all match the visible footage. The concept is called noise-coded illumination and was presented on August 10 at SIGGRAPH 2025 in Vancouver, British Columbia.

Read more of this story at Slashdot.

Terraform Labs founder Do Kwon pleaded guilty in U.S. federal court to conspiracy to defraud and wire fraud over the $40 billion collapse of TerraUSD and Luna in 2022. Reuters reports: Kwon, 33, who co-founded Singapore-based Terraform Labs and developed the TerraUSD and Luna currencies, entered the plea at a court hearing in New York before U.S. District Judge Paul Engelmayer. He had pleaded not guilty in January to a nine-count indictment charging him with securities fraud, wire fraud, commodities fraud and money laundering conspiracy. Accused of misleading investors in 2021 about TerraUSD - a so-called stablecoin designed to maintain a value of $1 - Kwon pleaded guilty to the two counts under an agreement with the Manhattan U.S. Attorney's office, which brought the charges. He faces up to 25 years in prison when Engelmayer sentences him on December 11, though prosecutor Kimberly Ravener said the government had agreed to advocate for a prison term of no more than 12 years provided he accepts responsibility for his crimes. "I made false and misleading statements about why it regained its peg by failing to disclose a trading firm's role in restoring that peg," Kwon said in court. "What I did was wrong."

Read more of this story at Slashdot.

ole_timer shares a report from the New York Times: Investigators have uncovered evidence that Russia is at least partly responsible for a recent hack of the computer system that manages federal court documents, including highly sensitive records with information that could reveal sources and people charged with national security crimes, according to several people briefed on the breach. It is not clear what entity is responsible, whether an arm of Russian intelligence might be behind the intrusion or if other countries were also involved, which some of the people familiar with the matter described as a yearslong effort to infiltrate the system. Some of the searches included midlevel criminal cases in the New York City area and several other jurisdictions, with some cases involving people with Russian and Eastern European surnames. Administrators with the court system recently informed Justice Department officials, clerks and chief judges in federal courts that "persistent and sophisticated cyber threat actors have recently compromised sealed records," according to an internal department memo reviewed by The New York Times. The administrators also advised those officials to quickly remove the most sensitive documents from the system. "This remains an URGENT MATTER that requires immediate action," officials wrote, referring to guidance that the Justice Department had issued in early 2021 after the system was first infiltrated. Documents related to criminal activity with an overseas tie, across at least eight district courts, were initially believed to have been targeted. Last month, the chief judges of district courts across the country were quietly warned to move those kinds of cases off the regular document-management system, according to officials briefed on the request. They were initially told not to discuss the matter with other judges in their districts.

Read more of this story at Slashdot.

None under active exploit…yet

Microsoft’s August Patch Tuesday flaw-fixing festival addresses 111 problems in its products, a dozen of which are deemed critical, and one moderate-severity flaw that is listed as being publicly known.…

An anonymous reader quotes a report from NPR: Boston Public Library, one of the oldest and largest public library systems in the country, is launching a project this summer with OpenAI and Harvard Law School to make its trove of historically significant government documents more accessible to the public. The documents date back to the early 1800s and include oral histories, congressional reports and surveys of different industries and communities. "It really is an incredible repository of primary source materials covering the whole history of the United States as it has been expressed through government publications," said Jessica Chapel, the Boston Public Library's chief of digital and online services. Currently, members of the public who want to access these documents must show up in person. The project will enhance the metadata of each document and will enable users to search and cross-reference entire texts from anywhere in the world. Chapel said Boston Public Library plans to digitize 5,000 documents by the end of the year, and if all goes well, grow the project from there. Because of this historic collection's massive size and fragility, getting to this goal is a daunting process. Every item has to be run through a scanner by hand. It takes about an hour to do 300-400 pages. Harvard University said it could help. Researchers at the Harvard Law School Library's Institutional Data Initiative are working with libraries, museums and archives on a number of fronts, including training new AI models to help libraries enhance the searchability of their collections. AI companies help fund these efforts, and in return get to train their large language models on high-quality materials that are out of copyright and therefore less likely to lead to lawsuits. "Having information institutions like libraries involved in building a sustainable data ecosystem for AI is critical, because it not just improves the amount of data we have available, it improves the quality of the data and our understanding of what's in it," said Burton Davis, vice president of Microsoft's intellectual property group. [...] OpenAI is helping Boston Public Library cover such costs as scanning and project management. The tech company does not have exclusive rights to the digitized data.

Read more of this story at Slashdot.