Homeland Security Investigations agents hit the outdoor retailer with a controversial subpoena as part of a dragnet search for the identities of protesters who entered a Minnesota church in March. Did you buy a beanie from REI recently? The Department of Homeland Security might be looking for you. New court filings allege that Homeland Security Investigations agents subpoenaed the outdoor retailer in March, requesting the transaction information for "all persons" in the greater Minneapolis-St. Paul area who had purchased a specific type of dark green beanie from the company since 2024.

Taiwanese authorities have indicted nine people, including employees of Nvidia and Super Micro Computer, over their alleged involvement in the illegal export of artificial intelligence servers to China. Eight of those indicted, including one employee of Nvidia's Taiwan unit and two employees of Supermicro's Taiwan unit, were charged with breach of trust and document forgery in connection with the illegal export of high-end AI servers, prosecutors in the port city of Keelung said in a statement on Monday. An additional 56 servers were seized at Taiwan's border, the statement said. Three of the defendants have been charged with embezzlement. Semiconductor powerhouse Taiwan is the world's largest producer of advanced chips used in AI applications.

A new toolkit for attorneys in Massachusetts targets the technologies police use--and conceal--to build criminal cases, from facial recognition to AI-written police reports. The American Civil Liberties Union of Massachusetts says it's releasing an online toolkit this week for criminal defense attorneys designed to uncover whether police used surveillance technologies --facial recognition, automatic license plate readers, gunshot detection systems, and more--to build the cases against their clients in secret. The toolkit, which the group describes as the first of its kind, is built around handcrafted legal motions that, when granted by a judge, will force prosecutors to reveal whether surveillance technology was used against a defendant. It covers an array of technologies, from "stingray" phone trackers and location data sold by commercial brokers to AI-drafted police reports and forensic tools that crack phones and siphon data from car infotainment systems. It also includes preservation motions--demands that surveillance data be saved before it is deleted automatically--aimed at government agencies and private vendors alike.

Elon Musk's xAI has filed a lawsuit against a South Carolina man who was arrested earlier this year on charges of sexually exploiting minors, alleging that he misused the company's AI tool to create sexually explicit content involving a child. The lawsuit, filed in a Texas federal court on Tuesday, alleges that Terry Harwood knowingly violated the company's terms of service to create the material. Despite expressly agreeing to abide by the xAI Terms of Service and Acceptable Use Policy, Defendant designed misleading prompts to circumvent Grok's built-in safeguards and then abused the tool to convert non-sexual photographs into sexually explicit images without the photograph subjects' knowledge or consent," the 12-page complaint said. The suit alleges that Harwood uploaded images of both adults and minors that were not sexual in nature and then tried to create so-called "deepfakes" that sexualised them. "A review of Defendant's xAI accounts further shows that on numerous occasions during the Relevant Period, Defendant submitted prompts to Grok to alter such images to sexualize the subjects of the images, which Grok responded to by refusing to follow the prompts on the basis that such material violated Grok's content moderation guardrails," the complaint added. "In response, Defendant repeatedly submitted further prompts, with alterations, in an effort to circumvent Grok's moderation efforts." The suit is the first brought by an AI company against one of its users and comes amid intense scrutiny of xAI around the world for allowing users to create this kind of content on the platform in the first place. Grok has been under the spotlight in Washington, been in the crosshairs of European regulators, and faced bans in both Malaysia and Indonesia regarding sexually explicit content that can be created on the platform. Earlier this year, Musk pushed back on allegations that Grok produced AI-generated sexualised images of children, especially nude images. "I [am] not aware of any naked underage images generated by Grok.

Things to Do in L.A. Tap to enable a layout that focuses on the article. This is read by an automated voice. Please report any issues or inconsistencies here . See more from the L.A. Times in Google Search. Federal prosecutors allege a 29-year-old Uber driver ignited the Lachman blaze that later became the Palisades fire, killing 12 people, leveling thousands of homes and causing billions in damage.

Japanese scientist Makoto Kuroda faces felony charges after allegedly poisoning a colleague's water bottle with toxic chemicals at UW-Madison's Influenza Research Institute.

Federal prosecutors charged three men linked to Supermicro with allegedly smuggling $2.5 billion in U.S. AI technology to China using fake documents, dummy servers, and shell companies.

The Equalized Odds (for short, EO) is one of the most popular measures of discrimination used in the supervised learning setting. It ascertains fairness through the balance of the misclassification rates (false positive and negative) across the protected groups -- e.g., in the context of law enforcement, an African-American defendant who would not commit a future crime will have an equal opportunity of being released, compared to a non-recidivating Caucasian defendant. Despite this noble goal, it has been acknowledged in the literature that statistical tests based on the EO are oblivious to the underlying causal mechanisms that generated the disparity in the first place (Hardt et al. 2016). This leads to a critical disconnect between statistical measures readable from the data and the meaning of discrimination in the legal system, where compelling evidence that the observed disparity is tied to a specific causal process deemed unfair by society is required to characterize discrimination. The goal of this paper is to develop a principled approach to connect the statistical disparities characterized by the EO and the underlying, elusive, and frequently unobserved, causal mechanisms that generated such inequality. We start by introducing a new family of counterfactual measures that allows one to explain the misclassification disparities in terms of the underlying mechanisms in an arbitrary, non-parametric structural causal model. This will, in turn, allow legal and data analysts to interpret currently deployed classifiers through causal lens, linking the statistical disparities found in the data to the corresponding causal processes. Leveraging the new family of counterfactual measures, we develop a learning procedure to construct a classifier that is statistically efficient, interpretable, and compatible with the basic human intuition of fairness. We demonstrate our results through experiments in both real (COMPAS) and synthetic datasets.

The Equalized Odds (for short, EO) is one of the most popular measures of discrimination used in the supervised learning setting. It ascertains fairness through the balance of the misclassification rates (false positive and negative) across the protected groups - e.g., in the context of law enforcement, an African-American defendant who would not commit a future crime will have an equal opportunity of being released, compared to a non-recidivating Caucasian defendant. Despite this noble goal, it has been acknowledged in the literature that statistical tests based on the EO are oblivious to the underlying causal mechanisms that generated the disparity in the first place (Hardt et al. 2016). This leads to a critical disconnect between statistical measures readable from the data and the meaning of discrimination in the legal system, where compelling evidence that the observed disparity is tied to a specific causal process deemed unfair by society is required to characterize discrimination. The goal of this paper is to develop a principled approach to connect the statistical disparities characterized by the EO and the underlying, elusive, and frequently unobserved, causal mechanisms that generated such inequality. We start by introducing a new family of counterfactual measures that allows one to explain the misclassification disparities in terms of the underlying mechanisms in an arbitrary, non-parametric structural causal model. This will, in turn, allow legal and data analysts to interpret currently deployed classifiers through causal lens, linking the statistical disparities found in the data to the corresponding causal processes. Leveraging the new family of counterfactual measures, we develop a learning procedure to construct a classifier that is statistically efficient, interpretable, and compatible with the basic human intuition of fairness. We demonstrate our results through experiments in both real (COMPAS) and synthetic datasets.

Despitethis noble goal, it has been acknowledged in the literature that statistical tests based ontheEOareoblivious totheunderlying causal mechanisms thatgenerated the disparity in the first place (Hardt et al. 2016).