PCWorld reports that OpenAI and Anthropic have temporarily relaxed usage limits for their latest AI models, including GPT-5.6 Sol and Claude's Fable 5. ChatGPT Plus, Business, and Pro subscribers can now access GPT-5.6 Sol without the previous five-hour usage window restrictions. This increased access appears driven by competition between the AI companies, giving users more opportunities to experience advanced reasoning capabilities. If you haven't kicked the tires yet on the latest and greatest ChatGPT and Claude models, this is your lucky week. OpenAI is (fittingly) opening the flood gates to GPT-5.6 Sol, its just-released and most powerful model, announcing Sunday that it's "temporarily" lifting the five-hour usage window for ChatGPT Plus, Business, and Pro subscribers. At the same time, Anthropic is - again - extending the trial period for Fable 5, its own new top-of-the-line, giving Claude subscribers another week of in-plan access .

PCWorld reports the new ChatGPT app launches for Mac with Windows version coming soon, featuring GPT-5.6 models and enhanced capabilities. The superapp introduces ChatGPT Work agent for complex projects, browser tabs, web publishing through'Sites' feature, and PC control functionality. GPT-5.6 offers three performance tiers (Sol, Terra, Luna) with access levels varying by subscription plan, targeting Claude Desktop's market position.

OpenAI says GPT-5.6 Sol will launch with a limited preview for trusted partners approved by the Trump administration before broader ChatGPT and API availability.

OpenAI has released its most powerful GPT-5.6 models (Sol, Terra, and Luna) but only to trusted partners and organizations following a U.S. government request for limited rollout. PCWorld reports that the flagship Sol model directly competes with Anthropic's restricted Fable and Mythos models, featuring enhanced layered safeguards against misuse. A wider public release remains uncertain due to regulatory concerns about the models' advanced capabilities and potential security implications. The U.S. government ban on Anthropic's powerful Fable and Mythos models has everybody spooked. That's why it's a little surprising that OpenAI, Anthropic's biggest rival, is slowly rolling out its latest and most powerful GPT models.

OpenAI Has New AI Models. The White House asked OpenAI to delay the rollout of its GPT-5.6 AI models, two weeks after Anthropic had to take its most advanced AI models offline. OpenAI is delaying the public release of its next generation of AI models, GPT-5.6, at the request of Trump's White House, the company confirmed on Friday. OpenAI said it would first share the models with a small set of customers, which will be preapproved by the US government. It will then work with the administration to slowly expand access.

OpenAI updated its GPT-5.5 Instant model with enhanced intent understanding and better ability to follow complex instructions and user clarifications. PCWorld reports the update includes improved location data usage for more effective local business and product searches. This minor release focuses on making AI interactions more intuitive, though its real-world impact requires further observation. Earlier this week, OpenAI announced via release notes that it has updated its most widely used AI model: GPT-5.5 Instant. According to the company, GPT-5.5 Instant should now be better at understanding the underlying intent of a question and keeping track of context across multiple messages.

GPT-5.5 Instant is now more capable at processing complex questions. OpenAI has updated GPT-5.5 Instant, the model you interact with the most when you use ChatGPT, to be better at understanding context and adapting to queries as you alter them to add more conditions or clarifications. The company updated ChatGPT's default model to GPT-5.5 Instant in May. Back then, it said that the model produced 52.5 percent fewer hallucinated statements during testing and 37.3 percent fewer factual errors. Now, the model has been upgraded to be more capable when it comes to identifying the underlying goal of a task or a question and carrying context over across multiple back-and-forths as you talk to it.

OpenAI Launches Full-Scale Effort to Patch Open-Source Bugs as It Takes on Anthropic's Mythos Amid concerns about AI models' cybersecurity capabilities, OpenAI revealed an improved version of GPT-5.5-Cyber and its "Patch the Planet" initiative to fix open-source software bugs. As fears about AI hacking capabilities grow, OpenAI on Monday made a slew of cybersecurity-focused announcements, including an improved version of its limited-access security-specialized model GPT-5.5-Cyber, As advances across the AI industry leave critical open-source projects at increasing risk of falling behind, though, the company also said on Monday that it is launching an effort known as Patch the Planet, founded with the prominent research-focused security firm Trail of Bits and in collaboration with vulnerability management firms HackerOne and Calif. The project has already begun its work offering free security consulting services to open source maintainers to not only help them find and patch vulnerabilities, but also support them in strengthening their code bases and incorporating AI security tools into their development process. The idea is to give individualized support to as many open-source projects as possible to improve both their current security and long-term resilience in a way that will actually be sustainable.

Large language models (LLMs) are increasingly used as judges for open-ended generation, as large-scale human evaluation is often expensive and difficult to scale, yet their preferences remain imperfect proxies for human judgment. Existing auditing pipelines often assume that a reliable subset of examples or clean supervision signals are available beforehand, for example from human annotation, heuristic filtering, or the outputs of strong judges. In LLM evaluation, this assumption is fragile: the initial split may inherit judge bias, while human verification is typically too scarce to define stable groups at scale. We propose AURA, an adaptive uncertainty--aware refinement framework for auditing pairwise LLM--as--a--judge decisions under selected human verification. AURA iteratively learns a human-consistency signal, propagates reliable evidence, and prioritizes uncertain comparisons for human review. The key idea is to treat trust in a judge as a latent quantity that is progressively refined as evidence accumulates. We provide a compact formulation, a stable refinement procedure, and a comprehensive evaluation on both synthetic and real pairwise LLM-answer data.

Multimodal large language models (MLLMs) remain vulnerable to transfer-based targeted attacks, where perturbations optimized on open-source surrogate encoders can generalize to closed-source MLLMs. A key challenge for improving adversarial transferability is to effectively capture the intrinsic visual focus shared across different models, such that perturbations align with transferable semantic cues rather than surrogate-specific behaviors. However, existing methods suffer from spatial-domain feature redundancy and surrogate-specific gradient signals, thereby hindering cross-model transferability. In this paper, we propose FRA-Attack, which addresses both challenges from a unified frequency-domain regularization perspective. For feature alignment, a high-pass DCT objective on patch features suppresses redundant global structures and concentrates the loss on the high-frequency band that carries the MLLMs' intrinsic visual focus. For gradient optimization, we introduce Frequency-domain Gradient Regularization (FGR), a \textit{model-agnostic} low-pass regularizer that modulates the surrogate gradient using only the geometric frequency coordinate, \textit{i.e.}, no surrogate-derived statistic is involved, so that FGR is model-agnostic by construction, removing surrogate-specific high-frequency artifacts while preserving transferable low-frequency directions. Together, the two components form a unified frequency-domain treatment of transferability. Extensive experiments on $15$ flagship MLLMs across $7$ vendors show that FRA-Attack achieves superior cross-model transferability, particularly with state-of-the-art performance on GPT-5.4, Claude-Opus-4.6 and Gemini-3-flash.