Model

• The Trump administration has initiated a regulatory crackdown on leading U.S. AI developers, specifically targeting Anthropic's advanced models. • This shift reverses previous government efforts to accelerate AI development by limiting regulatory hurdles for companies like Anthropic and Google. • Experts warn that these restrictions may inadvertently benefit Chinese model makers, allowing them to close the technological gap with the U.S.

• Meituan unveiled LongCat-2.0, a new large language model that the company claims performs comparably to Google's Gemini 3.1 Pro released in February. • The model was trained using domestic Chinese chips, marking a significant step in the country's effort to reduce reliance on U.S. hardware. • This development is a direct response to Washington's security-based restrictions on advanced chip exports to China.

• Chinese AI model GLM-5.2 is gaining rapid developer adoption after outperforming other open-source models and approaching Anthropic’s Opus 4.8 on a key agentic AI benchmark. • The model is specifically designed for enterprise tasks including coding, planning, and automation, operating at approximately one-fifth the cost of its competitors. • This development sparks a debate among experts regarding whether low-cost Chinese AI could disrupt the market dominance of US firms like OpenAI and Anthropic.

• New state-of-the-art stellar evolution models suggest that Earth may avoid being engulfed when the sun eventually dies. • While the sun is expected to expand to hundreds of times its current size during its red giant phase, these models hint that the planet might escape this fate. • This research is significant as it challenges previous assumptions about the inevitable destruction of the innermost planets.

• Financial regulators are increasing their scrutiny of artificial intelligence as the development of more powerful models creates new operational risks. • Key concerns center on a lack of transparency and visibility into advanced AI systems, making it difficult for authorities to maintain effective oversight. • A recent BBC report highlights a growing gap between the rapid pace of AI sophistication and the ability of existing monitoring mechanisms to control them.

• OpenAI has appointed former Uber India chief Prabhjeet Singh as Managing Director for India, which serves as the company's second-largest market. • The move comes as India sees significant adoption of OpenAI's technology, with over 100 million weekly ChatGPT users. • In a separate strategic shift, OpenAI has restricted the rollout of its GPT-5.6 models to selected partners following requests from the US government.

• Bernstein warns that India risks permanent dependence on foreign AI providers unless it develops its own sovereign large language models (LLMs) similar to DeepSeek. • The report argues that focusing solely on AI applications is insufficient, as control over foundational models is essential for long-term national competitiveness. • Current concerns highlight that India's AI Mission spreads funding too thin across compute, research, and startups, leaving limited resources specifically for foundational model development.

• Chinese AI developers are rapidly narrowing the technological gap with the U.S., specifically with the emergence of the GLM-5.2 model. • The new model demonstrates advanced capabilities in cybersecurity, challenging the previous dominance of American AI systems in this critical sector. • This shift is expected to impact global AI stock valuations, pricing power for tech firms, and the trajectory of international regulatory frameworks.

• Bihar CM Samrat Choudhary visited Miller High School in Patna to inspect its development as a model high school and interact with students. • During the visit, the CM viewed a demonstration of an AI robot developed by student Aditya and his team, which utilizes a custom Large Language Model (LLM). • The event highlights a growing trend of technological innovation among students in Patna and underscores the government's push for modernizing educational infrastructure.

• The robotics industry is shifting from experimental demonstrations toward real-world deployment, driven by advancements in safety and next-generation AI models. • World model AI lab startup Odyssey announced it has secured $310 million in Series B funding this week to further develop AI models for robotics. • This surge in investment underscores a growing confidence in the scalability of AI-powered robots and their ability to operate in complex environments.

• OpenAI has deferred the public rollout of its GPT-5.6 model following requests from the U.S. government for early access to frontier AI systems. • The delay is driven by increasing concerns in Washington regarding the national security risks associated with the deployment of highly powerful AI models. • This move highlights the growing tension between rapid commercial AI releases and the need for government oversight to prevent potential misuse.

• Hangzhou-based AI company DeepSeek plans to double its workforce following a multi-billion-dollar fundraising round. • The firm is immediately recruiting for 27 different roles, including development engineers, data engineers, AI product managers, and corporate staff in legal and finance. • This expansion aims to strengthen DeepSeek's ability to compete against major U.S. generative AI rivals such as OpenAI and Anthropic.

• New AI models are increasing the speed and complexity of cyberattacks while lowering the entry barriers for malicious actors. • Cybersecurity experts are currently engaged in a "cat-and-mouse" game to develop defenses that can keep pace with these evolving AI-driven threats. • The Five Eyes intelligence alliance—comprising Australia, Canada, New Zealand, the UK, and the US—continues to collaborate on gathering intelligence to respond to such security risks.

• OpenAI is launching a preview version of a more capable new AI model to a select group of partners before a broader public release. • The decision to stagger the rollout follows direct pressure from the Trump administration to restrict the immediate availability of the technology. • This move highlights the increasing influence of the U.S. government on the deployment timelines of cutting-edge artificial intelligence.

• The Trump administration has reportedly requested that OpenAI limit the launch of its next model, GPT-5.6, to government-approved partners. • This move follows President Trump's recent decision to scrap a previous executive order that established a voluntary framework for reviewing advanced AI models. • The shift suggests a transition toward a more consistent, government-led process for evaluating the most capable AI systems before they are released to the general public.

Sam Altman announces limited preview of GPT 5.6 in move that echoes launch of Anthropic’s MythosBusiness live – latest updatesOpenAI is staggering the release of its latest AI model after a request from the US government, in a move echoing the launch of Anthropic’s Mythos product.Sam Altman, the chief executive of the company behind ChatGPT, told staff this week that GPT 5.6 would be released in a limited preview to a small group of partners, according to the tech publication The Information. Continue reading...

• China’s Zhipu AI has released GLM-5.2, a cost-effective coding model designed to challenge global leaders in artificial intelligence. • The model marks a significant milestone as the first Chinese release to rank in the top three globally on a major industry benchmark. • This achievement is being described as a new "DeepSeek moment," signaling China's growing ability to produce high-performance models with lower computational costs.

• The "Five Eyes" intelligence alliance has warned that frontier AI models could significantly reshape the landscape of cyberattacks within months. • These advanced AI capabilities are expected to pose urgent security risks to governments, businesses, and nations like South Africa by 2026. • The shift matters because AI can automate and scale sophisticated cyber threats, potentially bypassing traditional security defenses more efficiently.

• The US government has urged Meta to voluntarily submit its AI models for federal review to identify and assess potential security vulnerabilities. • Meta is currently the only major US AI developer that has not reached an agreement to share its models with the government, following the April launch of its Muse Spark AI model. • This request is part of a broader effort by the Trump administration to intensify oversight and security monitoring within the rapidly evolving AI industry.

• Tang Jie, CEO of the leading Chinese AI startup Zhipu, publicly corrected Elon Musk in a rare social media exchange regarding the global AI race. • Tang disputed Musk's assertion that China is months behind the United States in developing frontier AI models, insisting that the gap is not as wide as suggested. • This clash highlights the intense geopolitical competition between the U.S. and China to achieve dominance in artificial intelligence and large-scale model development.

• A JP Morgan report reveals that while the US still dominates the AI landscape, China is rapidly closing the gap through the development of highly cost-efficient models. • Chinese firms such as DeepSeek, MiniMax, Xiaomi, and Alibaba are leading the "efficient frontier," attracting enterprises seeking to lower operating costs as US token pricing rises. • Data from OpenRouter indicates a surge in API calls to Chinese models, reflecting growing global market acceptance of these alternatives.

• A JP Morgan and ETEnterpriseai report indicates that while the US remains the global leader in AI, China is rapidly closing the gap by offering significantly cheaper operational costs for corporate environments. • US-developed frontier AI models continue to hold a technological edge, remaining essential for high-level tasks including cybersecurity, scientific discovery, and large-scale reasoning. • This competition highlights a strategic divide where China competes on affordability and accessibility, while the US maintains dominance in advanced, agentic systems.

• The "Five Eyes" intelligence alliance—comprising the U.S., U.K., Australia, New Zealand, and Canada—issued a rare joint statement warning that powerful AI models pose severe cybersecurity risks. • Officials state that generative AI could significantly lower the barrier to entry for malicious hackers while increasing the speed and complexity of attacks against governments and businesses. • This warning highlights a critical shift in the threat landscape, as AI enables attackers to automate sophisticated exploits that were previously too difficult for most actors to execute.

Aemo modelling shows electricity use will nearly double by 2050. Follow today’s news liveGet our breaking news email, free app or daily news podcastGood morning and welcome to our live politics blog. I’m Martin Farrer with the top overnight stories and then it will be Krishani Dhanji with the main action.There will be plenty of politics to come but we start with good news: thanks to the recent boom in batteries connected to the national power grid, we’re going to save money on new transmission lines. That’s despite our power use being expected to nearly double in the next decade or so. Continue reading...