The defendant used Grok to create sexualized images of adults and children, xAI says. SpaceXAI (as xAI) has filed a lawsuit against Terry Wayne Harwood, a 67-year-old man from South Carolina whom the company has accused of using Grok to generate sexual images of real people without their consent. In the complaint that xAI filed in Texas, it said that the defendant uploaded non-sexual images of numerous adults and minors to his two xAI accounts from December 8, 2025 until February 18, 2026. He then asked Grok to alter the photos or to create new images and videos depicting the adults and children in them in a pornographic manner or otherwise sexualizing them. The company said that Grok refused to follow his prompts on numerous occasions but that he repeatedly submitted edited prompts to circumvent the AI's safeguards. In one example mentioned in the filing, xAI said Harwood uploaded the photo of a fully dressed girl around 10 to 11 years old and then asked Grok to remove all of her clothing and make her do a Playboy model impression as she laid in bed.

A facial recognition system has been introduced at ticket gates at Ikebukuro Station on the Tobu Tojo Line in Tokyo. Hitachi and Tobu Railway on Wednesday introduced a facial recognition system at ticket gates at Ikebukuro Station in Tokyo's Toshima Ward. Passengers can pass through the ticket gates hands-free if they register their facial data and commuter pass information in advance. The companies hope to encourage other railway operators to adopt the system, which can be installed by simply adding cameras and other equipment to existing ticket gates. The system was jointly developed in cooperation with three ticket gate manufacturers, including Toshiba. This is the first time that the system has been installed at a major terminal station in Tokyo.

Since WIRED reported on Meta's NameTag face recognition system, company executives have made confusing and conflicting remarks about its very existence. Does a software feature exist if its code has been deployed to the devices of millions of people but they can't use it yet? Not if you work at Meta . The company's executives have spent the last few weeks making this semantic argument about NameTag, the in-development face-recognition system that Meta built for its smart glasses . The inevitable result is confusion, but that's easy enough to clear up.

Customers inside a B&M store, which is one of more than 100 businesses that will be using the technology. Customers inside a B&M store, which is one of more than 100 businesses that will be using the technology. Civil liberties groups say Facewatch system in stores such as Sainsbury's and B&M is'dangerous escalation' Fri 10 Jul 2026 06.19 EDTLast modified on Fri 10 Jul 2026 06.57 Facial recognition technology in shops will soon alert police in real time to the presence of serious offenders, with civil liberties groups warning of a "dangerous escalation" towards surveillance and criminalisation in the retail sector. Facewatch, a facial recognition system used by more than 100 businesses including Sainsbury's, B&M and Spar to monitor thieves, said it was launching a UK-first feature to "alert police instantly when the most serious offenders trigger a live facial recognition match".

Bees seem to show when they are pleased and like something, rather than just needing it, in one of the strongest signs yet that insects have subjective experiences. In recent decades, it has become clear that bees are capable of more complex behaviours than we previously thought, such as counting and demonstrating a sense of rhythm . But discerning whether they have inner states akin to our emotions is more difficult. For one thing, insects don't have the flexible facial musculature of mammals, which we use to communicate our feelings. "How can we get any behavioural readout of these insects with a hard body and their mask of a face," asks Andrew Barron at Macquarie University in Sydney, Australia.

Visual prompting (VP) has emerged as a promising parameter-efficient fine-tuning approach for adapting pre-trained vision models to downstream tasks without modifying model parameters. Despite offering advantages like negligible computational overhead and compatibility with black-box models, conventional VP methods typically achieve lower accuracy than other adaptation approaches. Our analysis reveals two critical limitations: the restricted expressivity of simple additive transformation and a tendency toward overfitting when the parameter count increases. To address these challenges, we propose ACAVP (Affine, Color, and Additive Visual Prompting), which enhances VP's expressive power by introducing complementary transformation operations: affine transformation for creating task-specific prompt regions while preserving original image information, and color transformation for emphasizing task-relevant visual features. Additionally, we identify that overfitting is a critical issue in VP training and introduce TrivialAugment as an effective data augmentation, which not only benefits our approach but also significantly improves existing VP methods, with performance gains of up to 12 percentage points on certain datasets. This demonstrates that appropriate data augmentation is universally beneficial for VP training. Extensive experiments across twelve diverse image classification datasets with two different model architectures demonstrate that ACAVP achieves state-of-the-art accuracy among VP methods, surpasses linear probing in average accuracy, and exhibits superior robustness to distribution shifts, all while maintaining minimal computational overhead during inference. Our code is available at https://github.com/s-enmt/ACAVP.

We introduce Eagle2.5, a frontier vision-language model (VLM) for long-context multimodal learning. Our work addresses the challenges in long video comprehension and high-resolution image understanding, introducing a generalist framework for both tasks. The proposed training framework incorporates Automatic Degrade Sampling and Image Area Preservation, two techniques that preserve contextual integrity and visual details. The framework also includes numerous efficiency optimizations in the pipeline for long-context data training. Finally, we propose Eagle-Video-110K, a novel dataset that integrates both story-level and clip-level annotations, facilitating long-video understanding. Eagle2.5 demonstrates substantial improvements on long-context multimodal benchmarks, providing a robust solution to the limitations of existing VLMs.

What Happened to Your Face? How the human countenance became something to study, edit, optimize, and scan. The physiognomists promised that your character could be read from your features. Certain forms of facial-recognition technology have revived that old fantasy in digital form. Several months ago, my partner and I bought an apartment in South London. Our previous home was a rental in which, for reasons best known to the landlord, there were mirrors everywhere. The bathroom had two; there was one outside on the terrace; in the bedroom, mirrored panels stretched across a twenty-foot-long wall. On moving day, we realized that we had a problem: the new apartment was mirror-free, and because we'd been so spoiled we weren't bringing one of our own. We spent a few days filling our drafty rooms, decanting books, building furniture, and dressing every morning without seeing ourselves in profile. It was a couple of weeks before we bought a simple mirror, wooden and round, to hang above the bathroom sink. By then, I joked, we didn't recognize ourselves.

Camera, wifi and design updates bring welcome upgrades to Ring's top model in wired or battery flavour The Guardian's journalism is independent. We will earn a commission if you buy something through an affiliate link. R ing's recent revamp of its popular video doorbells with a more modern design is led by the top-of-the-line Video Doorbell Pro 3, which gains much-needed upgrades with a 4K camera and better wifi plus new interesting AI features. The Guardian's journalism is independent. We will earn a commission if you buy something through an affiliate link.

We introduce ROGR, a novel approach that reconstructs a relightable 3D model of an object captured from multiple views, driven by a generative relighting model that simulates the effects of placing the object under novel environment illuminations. Our method samples the appearance of the object under multiple lighting environments, creating a dataset that is used to train a lighting-conditioned Neural Radiance Field (NeRF) that outputs the object's appearance under any input environmental lighting. The lighting-conditioned NeRF uses a novel dual-branch architecture to encode the general lighting effects and specularities separately. The optimized lighting-conditioned NeRF enables efficient feed-forward relighting under arbitrary environment maps without requiring per-illumination optimization or light transport simulation. We evaluate our approach on the established TensoIR and Stanford-ORB datasets, where it improves upon the state-of-the-art on most metrics, and showcase our approach on real-world object captures.