AI & Robotics News: WeRide’s Saudi Robotaxi Launch, Waymo in Ontario, and Mystery ‘Horizon Alpha’ AI
WeRide Secures Saudi Arabia’s First-Ever Robotaxi Permit
WeRide has officially become the first company granted a permit for commercial robotaxi operations in Saudi Arabia. This significant milestone marks the sixth country where WeRide has received autonomous vehicle (AV) approvals, joining China, the UAE, Singapore, France, and the US. In a strategic collaboration with Uber and Ai Driver, WeRide has already initiated pilot robotaxi operations in Riyadh. The company aims for a full commercial service launch by the end of 2025. This permit was awarded following the successful completion of Saudi Arabia’s rigorous AV Regulatory Sandbox program.
Sources:
Waymo Lobbies Ontario for Favorable Autonomous Vehicle Regulations
Google’s autonomous vehicle unit, Waymo, is actively engaging with the Ontario government to help shape the province’s future regulatory framework for driverless vehicles. Signaling a strong interest in commercial deployment in Canada, consultants from StrategyCorp Inc. have registered to lobby on Waymo’s behalf. Having already launched successful driverless taxi services in several U.S. cities, Waymo is now advocating for regulations that would permit its autonomous technology to serve Canadians. Transport Canada has confirmed its ongoing engagement with Waymo to better understand its testing and deployment protocols as it develops new guidance for the autonomous vehicle industry.
Sources:
Mystery ‘Horizon Alpha’ AI Model with 256K Context Appears on OpenRouter
A new, unidentified generative AI model named ‘Horizon Alpha’ has appeared on the OpenRouter platform, sparking curiosity in the AI community. The model boasts a large 256K token context window and can process up to 128K input tokens. Speculation is rampant that ‘Horizon Alpha’ could be an early test version of OpenAI’s highly anticipated GPT-5 or a new open-source model, though this is unconfirmed. Initial user testing indicates strong performance as a conversational assistant, but it may lag behind established models like GPT-4.1 in coding capabilities. Currently in a free beta, the model’s provider is collecting data to refine its performance ahead of an official launch with comprehensive benchmark data.
Sources:
Experts Warn: Generative AI Poses New Child Safety & CSAM Threats
Child safety experts are sounding the alarm over new threats posed by generative AI, which is being used to accelerate the creation and distribution of child sex abuse material (CSAM). According to the Internet Watch Foundation (IWF), malicious actors are using fine-tuned AI models to generate new, synthetic abusive imagery of known child victims and even famous children. Furthermore, deepfake technology is being exploited to insert a child’s face into adult pornographic content. The National Center for Missing and Exploited Children (NCMEC) also warns that generative AI facilitates the creation of convincing fake social media profiles used to lure children. This has led to a disturbing increase in AI-generated CSAM found on the publicly searchable ‘clear web’.
Sources:
CrowdStrike Report: Cloud & Identity-Based Cyberattacks Skyrocket
A new report from cybersecurity leader CrowdStrike reveals a dramatic escalation in the sophistication of cyberattacks, highlighting a sharp increase in cloud-focused and identity-driven intrusions. The 2025 Threat Hunting report shows a staggering 136% rise in cloud intrusions in the first half of 2025 compared to all of 2024. Threat actors are employing advanced tactics, exploiting cloud misconfigurations and abusing control planes for lateral movement and persistence. The report also notes a 185% spike in cyberattacks targeting government entities. Notably, adversary groups like Scattered Spider have drastically accelerated their attack timelines, in some cases moving from initial compromise to ransomware deployment in under 24 hours.
Sources: