AI News

AI News Brief: OpenAI launched a brand-new AI Agent called ‘ChatGPT Work’, focused on taking action directly across applications and files. Instead of just answering questions or giving suggestions, it actually executes tasks and delivers real results. The core design lies in its ability to sustain long-duration work — for complex projects, it can run continuously for several hours, maintaining a full grasp of the goal throughout, without requiring users to repeatedly step in. You just set a goal, and ChatGPT Work breaks down the task steps on its own, operates the relevant tools and files, and ultimately turns the goal into completed work. This marks a fundamental shift from ChatGPT’s traditional Q&A interaction model, signaling that OpenAI is officially pushing its flagship product into the era of agentic workflows. However, the original source provides limited technical detail and concrete app support — refer to the original link for the full picture.

AI News Brief: SpaceXAI officially released its latest model Grok 4.5 on Wednesday — it’s also the first large language model the company has launched since going public. The company positions Grok 4.5 as an all-purpose workhorse model focused on automating everyday knowledge work including programming, office documents, research, and content writing. On cost competitiveness, SpaceXAI claims Grok 4.5’s t…

AI News Brief: Taiwan’s exports are hitting record highs, driven by the global AI wave. As a key global chip supplier, Taiwan’s trade performance keeps breaking records. According to the headline, first-half 2026 exports grew nearly 50% year-over-year, fueled mainly by explosive demand for AI-related semiconductors and hardware. Keelung Port’s cargo volume also mirrors this export boom. Thanks to strong exports, Taiwan’s economic growth is projected to hit its fastest pace since 2010…

AI news brief: Saudi Arabia’s LEAP tech show made its East Asian debut in Hong Kong, drawing over 450 exhibitors from mainland China, Hong Kong, and Saudi Arabia across biotech, robotics, and software. Chinese humanoid robots drew crowds with dance routines, some draped in shemagh headscarves to appeal to Gulf visitors. This reflects Hong Kong’s strategic pivot toward the Middle East under rising Western geopolitical pressure, with the most concrete result being a late-2024 MoU with Saudi sovereign wealth fund PIF to establish a $1 billion fund supporting Hong Kong businesses entering the Saudi market.

AI News Brief: Google’s SynthID watermark system successfully helped expose a high-profile AI-generated fake image this week, becoming a landmark case in deepfake prevention technology.

AI News Brief: The latest Fed meeting minutes show FOMC members divided on whether to raise rates, but most view strong AI infrastructure demand as a major source of inflation pressure. The minutes note that sustained massive AI-related demand is ’likely to continue putting upward pressure on tech product and electricity prices’ — a phenomenon dubbed ‘chipflation’, driven by data centers snapping up high-end semiconductors and players scrambling for power, which in turn pushes up end prices for consumer electronics and energy.

AI news flash: a closed blockchain protocol is not a safe one. On June 14 this year, hackers exploited a smart contract vulnerability to steal $2.1 million from the Aztec Connect protocol, which had been shut down since March 2023; five days later, the smart contract of the decentralized exchange mySwap was attacked for a $300,000 loss, even though the platform had stopped accepting new liquidity for over six months. Abandoned contracts are still attack targets—risk doesn’t disappear just because a protocol shuts down…

AI Briefing: The Aspire team built a cross-repository documentation automation pipeline using GitHub Agentic Workflows. After product code is merged, the system automatically detects the scope of changes and generates a corresponding draft doc pull request in the docs repository for SMEs to review and confirm, instead of relying on engineers to write documentation manually. This design runs doc updates in parallel with the development workflow, shortening the window where the feature is live but docs haven’t synced yet.

AI News Brief: OpenAI recently published an in-depth analysis report on SWE-Bench Pro, raising several concerns about this widely-adopted benchmark in the AI coding evaluation space. SWE-Bench Pro has long been regarded as an important indicator for measuring AI models’ ability to solve real software engineering problems, and many mainstream models cite it as the basis for their public performance claims. However, OpenAI’s analysis points out that the benchmark has notable issues in reliability and accuracy that could lead to misjudgments about actual model performance—in other words, scoring high on SWE-Bench Pro doesn’t necessarily reflect real-world coding capability. This analysis strikes at a core pain point in AI evaluation: when a benchmark itself is flawed or over-optimized, the industry’s progress metrics can lose their reference value.

AI News Brief: OpenAI recently publicly outlined its core principles and framework for partnerships with government and national security agencies. According to the summary, OpenAI has established three core guidelines for government and national security collaborations: responsible AI applications, democratic accountability mechanisms, and public safety guarantees. ‘Responsible AI applications’ means technology deployment must comply with ethical norms and must not be used in scenarios that violate human rights or enable abuse; ‘Democratic accountability’ emphasizes that even when the partner is a government security agency, transparency must be maintained and oversight from democratic institutions preserved, rather than allowing AI capabilities to override public interests; ‘Public safety’ reflects OpenAI’s bottom-line stance that its technology must not be used to harm civilians or public interests.