Back to School Mashable's Best: E-readers, robovacs, laptops, earbuds, smart home and more Mashable Selects Say More Look Up Trending Now Good Connection: Uplifting stories for a digital age Creator Playbook Switch Off Mashable Voices Safety Net Versus All Series Minnesota'nudification' law survives xAI's request for a legal injunction Minnesota has temporarily won the right to sue companies for up to $500,000 for every nudification enabled by artificial intelligence. The District Court of Minnesota has rejected the efforts of Elon Musk's xAI to seek an injunction to suspend the law until a court could hear its constitutional challenge to the law itself, according to a report by Reuters . Judge Donovan W. Frank, in his Conclusion, argued that xAI has not demonstrated irreparable harm and the balance of harms tips steeply in favor of the State. The bill Musk is trying to fight was recently passed into law by Minnesota, becoming the nation's first anti-nudification law . Companies found in violation of this law would be subject to fines of up to $500,000 every time artificial intelligence is used to alter images to depict the intimate parts of a person in such a way that a reasonable person would believe that the intimate part belongs to the identifiable individual.

Foreign recruitment and external support networks are sustaining and expanding the civil war in Sudan by providing weapons, military technology and fighters, to the detriment of civilians, according to a United Nations-backed investigation. A report by the UN Independent International Fact-Finding Mission for the Sudan, released on Thursday, said transnational entities operating across several countries recruited, trained and deployed foreign personnel to support the paramilitary Rapid Support Forces (RSF). Since April 2023, the conflict between the SAF and the RSF has killed tens of thousands of Sudanese and forced millions to flee their homes. "This support fed into RSF operations in Zamzam camp and later during the takeover of el-Fasher, where the [UN] Mission documented grave human rights and international humanitarian law violations and international crimes, including conduct displaying hallmarks of genocide," mission member Mona Rishmawi said. The report said external entities have substantially enhanced the drone capabilities of both the RSF and the SAF, enabling them to strike far beyond front lines.

Things to Do in L.A. Tap to enable a layout that focuses on the article. Traffic makes its way down Ocean Avenue in Santa Monica. The city is using AI-powered cameras to identify and cite motorists who block bike lanes. This is read by an automated voice. Please report any issues or inconsistencies here .

The European Commission claims that Google boosted its own apps and products to the top of search rankings to the detriment of its competitors. The European Commission has levied a $1 billion penalty against Google over alleged competition law violations. An EC investigation found that Google had abused its dominance in the European Union's search and app store markets to funnel people toward its own apps and services, in violation of the EU's Digital Markets Act . The body has ordered Google to refrain from giving preferential treatment to its own services--such as shopping, accommodations, transport, and flights--in search rankings. Google must also allow app developers to communicate and transact with users outside the Play Store, where it takes a commission on sales .

The company donated more than $3.6 million to conservative groups between 2020 and 2025, including $1 million to the MAGA Inc. super PAC. The lettuce supplier at the center of the turbo diarrhea outbreak has spent millions of dollars to sway sentiment around food regulation and elect Donald Trump and other MAGA Republicans, according to public filings. Taylor Farms donated more than $2 million to conservative political groups in 2025, according to Federal Election Commission filings. That includes a $1 million donation to MAGA Inc., the Trump-centric super PAC, and $1.1 million to other super PACs dedicated to electing Republicans. From 2020 through the end of 2024, the company donated more than $1.6 million to conservative PACs, including a total of $850,000 to AFP Action, an anti-regulation super PAC with ties to the Koch network.

Follow this section to personalize your feed and get instant alerts. Follow Go to your personalized feed WHY FOLLOW? Smart Alerts: Get notified about major news as it happens. Follow this tag to personalize your feed and get instant alerts. Follow Go to your personalized feed WHY FOLLOW?

Data fusion techniques integrate information from heterogeneous data sources to improve learning, generalization, and decision-making across data sciences. In causal inference, these methods leverage rich observational data to improve causal effect estimation, while maintaining the trustworthiness of randomized controlled trials. Existing approaches often relax the strong "no unobserved confounding" assumption by instead assuming exchangeability of counterfactual outcomes across data sources. However, when both assumptions simultaneously fail--a common scenario in practice--current methods cannot identify or estimate causal effects. We address this limitation by proposing a novel partial identification framework that enables researchers to answer key questions such as: Is the causal effect positive/negative? and How severe must assumption violations be to overturn this conclusion?

We propose a practical sequential test for auditing differential privacy guarantees of black-box mechanisms. The test processes streams of mechanisms' outputs providing anytime-valid inference while controlling Type I error, overcoming the fixed sample size limitation of previous batch auditing methods. Experiments show this test detects violations with sample sizes that are orders of magnitude smaller than existing methods, reducing this number from 50K to a few hundred examples, across diverse realistic mechanisms. Notably, it identifies DP-SGD privacy violations in under one training run, unlike prior methods needing full model training.

Deep learning models are increasingly deployed in safety-critical tasks where predictions must satisfy hard constraints, such as physical laws, fairness requirements, or safety limits. However, standard architectures lack built-in mechanisms to enforce such constraints, and existing approaches based on regularization or projection are often limited to simple constraints, computationally expensive, or lack feasibility guarantees. This paper proposes a model-agnostic framework for enforcing input-dependent linear equality and inequality constraints on neural network outputs. The architecture combines a task network trained for prediction accuracy with a safe network trained using decision rules from the stochastic and robust optimization literature to ensure feasibility across the entire input space. The final prediction is a convex combination of the two subnetworks, guaranteeing constraint satisfaction during both training and inference without iterative procedures or runtime optimization. We prove that the architecture is a universal approximator of constrained functions and derive computationally tractable formulations based on linear decision rules. Empirical results on benchmark regression tasks show that our method consistently satisfies constraints while maintaining competitive accuracy and low inference latency.

Integrating logical constraints into object detection models for autonomous driving (AD) is a promising way to enhance their compliance to rules and thus increase the safety of the system. In this, t-norms have been utilized to calculate the constrained loss, i.e., the violations of logical constraints as losses. While prior works have statically selected few t-norms, we conduct an extensive experimental study to identify the most effective choices, as suboptimal t-norms can lead to undesired model behavior. For this, we present MOD-ECL, a neurosymbolic framework that implements a wide range of t-norms and can use them in an adaptive manner, with an algorithm that selects well-performing t-norms during training and a scheduler that regulates the impact of the constrained loss. We evaluate its effectiveness on the ROAD-R and ROAD-Waymo-R datasets for object detection in AD with attached common-sense constraints. Our results show that careful selection of parameters is crucial for good behavior of the constrained loss and that our framework allows us to obtain not only lower constraint violation but in some cases also an increase in detection performance. Furthermore, our methods allow fine control over the tradeoff between accuracy and violation.1