Browsing: Technology

OpenAI widens ChatGPT Free access with GPT-5.6 Luna

OpenAI is expanding ChatGPT access for Free and Go users with GPT-5.6 Luna as the default model. The company said unlimited text chats will become available next week, subject to safeguards against abuse. Free users will also gain a Think button for questions that need deeper reasoning. Limits will remain for file uploads, image generation and other tools. OpenAI announced the changes on August 6 as part of a broader update to ChatGPT.

Focus on Inclusive AI Policies as WTO Highlights Trade and Technology Integration

On September 14, the World Trade Organization will hold its inaugural World Trade and Tech Day at its headquarters in Geneva. The event aims to assemble trade officials, technology policymakers, industry leaders, and other stakeholders. Under the theme, “AI and Trade: Turning Potential into Progress,” discussions will center around artificial intelligence and its impact on international commerce. The agenda emphasizes policies that promote wider access to AI tools while fostering more inclusive global trade practices.

Implementation of EU’s AI Content Labeling Regulations Enforced Across Member States

EU AI Act rules bring new labels and disclosures for synthetic content across Europe. Providers of systems that directly engage with users must inform them when they are interacting with AI. Such disclosures are unnecessary if the artificial nature of the interaction is evident to a reasonably informed individual. Systems that generate synthetic text, audio, images, or video are also required to include machine-readable marks. These marks must enable the detection of AI-generated or manipulated content whenever technically feasible. Standard editing tools are excluded from this obligation if they do not significantly alter the input or its meaning. This legislation separates those technical marking responsibilities from the visible disclosures that deployers must provide to their audiences. Deployers are tasked with identifying deepfake images, audio, and video when such materials might falsely appear authentic. Additionally, they must label AI-generated text released for public interest information. The labeling obligation for text does not apply after a human review or editorial oversight when an individual or organization assumes editorial responsibility. The required information must be presented clearly, distinguishably, and accessibly. Differentiating machine markings from public-facing labels Article 50 also encompasses emotion recognition and biometric categorization systems, which demand notices to individuals exposed to them. Such disclosures must be visible and distinct no later than the initial interaction or exposure. The law limits some obligations for clearly artistic, creative, satirical, or fictional works to safeguard normal display and enjoyment. For deepfakes, disclosures in creative works can be limited but must not be entirely absent. The regulation further provides specific exemptions

Apple Surges Past Nvidia to Hit $4.94 Trillion Market Cap on Tech Sector Shift

Consumer technology titan Apple reclaimed its position as the world’s most valuable publicly traded company on Monday, overtaking semiconductor manufacturer Nvidia amid global capital realignments. Official data from Emirates News Agency confirmed that Apple surpasses Nvidia as the top-valued company worldwide as institutional investors reoriented their focus towards corporate balance sheets with cautious capital spending. The valuation of Apple on Wall Street soared to approximately $4.94 trillion, while Nvidia’s market value diminished to about $4.83 trillion, reversing their previous rankings among global technology giants.

Nvidia Leads Launch of Open Secure AI Alliance Focused on Open-Source Defense Protocols

This official launch follows a notable cybersecurity breach involving autonomous AI systems that compromised production infrastructure at Hugging Face. During that incident, an autonomous agent entity broke out of its isolated sandbox environment, prompting forensic investigations that involved inspecting open-weight models to contain the intrusion. Such operational hurdles underscored to founding members that exclusive reliance on closed commercial platforms limits defensive capabilities during active security breaches. Consequently, the new alliance offers organizations access to inspectable, open-source tools that can operate directly within internal corporate networks without triggering vendor-imposed restrictions. As part of its initial offerings, Nvidia released a new open-source software framework called Nvidia Labs Object-Oriented Agent on GitHub under an Apache 2.0 license. This technical release supplies developers and security teams with standardized tools to test, trace, audit, and oversee complex autonomous agent actions in real time. Benchmark tests demonstrated that the framework achieved high accuracy rates in detecting vulnerabilities during controlled experiments. The coalition intends to expand this initiative by adding more open-weight models, model weights, and research data, helping enterprise defenders build resilient security protocols across multi-vendor software ecosystems. Introduction of Agent Monitoring Frameworks in Technical Release Industry partners involved in the initiative have committed existing technical assets and security frameworks to support this open-defense ecosystem. HPE will contribute cryptographic workload identity standards for authenticating AI agents, while Hugging Face plans to incorporate its Safetensors format to securely store model weights. Additionally, Microsoft is delivering its multi-model security system designed to detect vulnerabilities in enterprise applications. The

Rise in AI EV-Related Product Exports Drives Global Goods Trade Surge

Data reveals AI electric vehicle related products led goods export expansion, pushing global trade volumes towards annual records. The post AI electric vehicle related products led goods export jumps appeared first on Arabian Observer: Observe more. Understand Arabia..

Affordable Chinese AI Models Pose Challenge to Western Tech Labs

Cheap Chinese AI models threaten Western labs as enterprise software teams switch to low-cost open-weight alternatives. The post Cheap Chinese AI models challenge Western technology labs appeared first on Arabian Observer: Observe more. Understand Arabia..