

IFLYTEK Releases Edge AI Models Supporting Million-Token Context
iFLYTEK's wholly owned subsidiary recently released and open sourced Spark X2.5-4B and Spark X2.5-1.7B. These are presented as the first edge models to natively support up to one million tokens of context, according to the Chinese AI firm. This development aims to significantly expand the memory window for on device large models.
48 minutes ago


NYC Bans AI for K-8 Students, Most Restrictive US Policy
New York City officials have announced a one-year moratorium on the use of artificial intelligence by students in public elementary and middle schools. This measure represents a significant step by public bodies to establish clear guidelines concerning new technologies within educational environments. It is considered the most restrictive policy across the United States to date.
4 hours ago


Manus AI Resumes Independent Operations After Beijing Blocks Meta Deal
The AI start up Manus has formally recommenced its independent operational activities. This development follows the intervention by Beijing, which obstructed its USD 2 billion acquisition by Meta Platforms. This concludes a period of more than four months of significant uncertainty within the technology sector for the company.
19 hours ago


Amazon's Zoox, Waymo Expand Driverless Ride-Hailing Across US
According to Reuters, Amazon's autonomous vehicle subsidiary, Zoox, and Alphabet's Waymo, both prominent Big Tech entities, have separately announced expansions of their driverless ride hailing operations into further United States cities. These declarations follow the recent commencement of paid services by Zoox in one of its established operational locations. This development marks a period of intensified activity within the autonomous transport sector.
20 hours ago


OpenAI Denies Apple's Trade Secret Theft Claims, Cites No Evidence
OpenAI has denied allegations of trade secret theft from Apple, stating the iPhone maker provided no evidence of confidential information being stolen by former employees. According to Reuters, the AI organisation described the dispute as being "a mess of Apple's own making," in a filing in a US District Court in San Jose, California. This legal action escalates tensions between the two prominent technology firms.
22 hours ago


Singapore Strengthens Online Safety Measures for Young Users
According to The Straits Times, Big Tech social media platforms operating in Singapore will soon face new requirements to implement robust safeguards for teen users. These legislative changes, scheduled for early next year, aim to protect mental health through features like daily time limits, restrictions on infinite scrolling, autoplay functions, and direct messages from unknown individuals.
1 day ago


OpenAI Identifies Advanced Model Requiring Enhanced Safeguards
OpenAI, the artificial intelligence research organisation, has determined that one of its forthcoming models possesses capabilities necessitating additional safety protocols. These stringent measures are deemed essential for both the model's ongoing development and its prospective public introduction, according to officials within the company.
1 day ago


Meta Removes Malicious Sexually Explicit Ads After Indian Government Alert
Meta removed numerous advertisements from its Facebook and Instagram platforms after the Indian government issued an alert. The adverts reportedly used sexually explicit content to entice users into downloading malicious software. This software was designed to steal banking credentials and subsequently drain bank accounts, according to the alert.
2 days ago


StartLux Small AI Rivals Trillion-Parameter Systems in Key Test
StartLux, a Shanghai-based artificial intelligence organisation, achieved second place in the China Academy of Information and Communications Technology MCP test. Its 27 billion parameter local model surpassed DeepSeek-V4-Flash, a considerably larger model, performing at a level often associated with trillion parameter systems. This outcome suggests a notable capability for smaller AI architectures.
2 days ago


Zhipu AI’s GLM-5.3-Flash Achieves Top Global AI Call Volume
Zhipu AI's GLM-5.3-Flash, previously an anonymous entity known as Ox Alpha, has achieved the highest global AI model call volume. This marks 18 consecutive weeks of Chinese models leading worldwide inference volume. The recent surge saw global AI model traffic reach 113 trillion tokens, a significant weekly increase.
2 days ago























































