Seattle Public Schools is tightening restrictions on student technology use. YouTube will be blocked for kindergarten through fifth grade students starting Monday, while Minecraft will be blocked for all K-12 students except those using it in classroom lessons. The AI chatbot Gemini is now banned districtwide. Superintendent Ben Shuldiner says the district is rethinking pandemic-era technology policies, and starting in January, third through fifth graders will lose school-issued laptops with screen time capped at 60 minutes per day.
Google is drawing criticism for changes to its search engine for California K-12 students. Starting in January, students will be blocked from using Google Search unless their districts choose the AI-powered version, with no option for a traditional search experience. Parent advocate Lila Byock called the move an act of bullying, while Common Sense Media said removing Google search limits student learning opportunities. Some districts may struggle to find alternatives due to long contracts and training costs.
Google is also turning off AI tools for schoolchildren by default worldwide. Gemini, Gemini Notebook, Gemini in Classroom, and Search for K-12 student accounts will remain disabled until administrators enable them. New York Mayor Zohran Mamdani criticized Google for trying to force AI into public schools. More than 150 million students worldwide use Google Classroom.
Key Takeaways
- Seattle Public Schools blocks YouTube for K-5 students and bans Gemini AI districtwide
- Google forces AI-powered search on K-12 students in California with no traditional option
- Google turns off Gemini and Search tools by default for K-12 accounts worldwide
- Match Group bans AI bots from Tinder and Hinge to preserve human connection
- Go.AI launches Go1 MAX appliance serving up to 8,000 concurrent users for enterprises
- FINRA requires human review of all AI-generated content in financial communications
- Stacey Abrams calls for careful AI regulation and community input on data centers
- Apple removes waitlist for updated Siri, making new AI features available immediately
- UC Berkeley Law launches course on AI use in litigation and legal work
- Anthropic security test shows Claude contained by microsegmentation after initial breach
Seattle Public Schools limits YouTube, Minecraft, and AI access for students
Seattle Public Schools will block YouTube for kindergarten through fifth grade students starting Monday. Minecraft will be blocked for all students in kindergarten through 12th grade, except for classes that use it in lessons. The AI chatbot Gemini is now banned districtwide. Superintendent Ben Shuldiner says the district is rethinking pandemic-era technology policies. Starting in January, third through fifth graders will lose school-issued laptops and have screen time capped at 60 minutes per day.
Google faces backlash over AI search changes for California schools
Google announced that starting in January it will block K-12 students from using its search engine unless districts choose its AI-powered version. There is no option for a search engine without AI. Parent advocate Lila Byock called the move an act of bullying. Common Sense Media said removing Google search limits student learning opportunities. Some districts may struggle to find alternatives because of long contracts and training costs.
Google turns off AI tools for schoolchildren by default
Google said it will turn off Gemini, Gemini Notebook, Gemini in Classroom, and Search for K-12 student accounts until administrators enable them. The changes will roll out worldwide starting in the United States. Schools will not be able to offer Google Search without AI features. New York Mayor Zohran Mamdani criticized Google for trying to force AI into public schools. More than 150 million students worldwide use Google Classroom.
Dating app company bans AI bots to keep human connection
Match Group, the company behind Tinder and Hinge, has banned AI agents from its dating platforms. The company wants humans to review profiles and send messages instead of automated bots. The decision comes as tech companies rush to add AI to their services. Match Group believes some friction is needed to keep dating experiences authentic and meaningful. The company urges other online platforms to slow down and think carefully about adding AI agents.
Go.AI launches powerful on-premises AI appliance for organizations
Go.AI introduced the Go1 MAX, an on-premises AI appliance that can serve up to 8,000 concurrent users from a single device. It is designed for banks, hospitals, insurance companies, and large enterprises that need private AI. The system allows teams to train and serve AI models without cloud quotas or GPU queuing. Co-Founder and CEO David Moscatelli said the product gives regulated organizations control over their data and AI infrastructure. Go.AI was formerly known as Go Abacus Corporation.
FINRA rules require human review for all AI content
FINRA states that artificial intelligence tools must follow the same compliance rules as human writers. Financial firms must have a person make the final decision on every communication before it is sent. This creates a challenge because AI can create content much faster than teams can review it. The regulator also warned that social media posts are hard to control once they are online. Companies like Red Oak are using tools to help manage these reviews while keeping human oversight.
Stacey Abrams discusses AI risks and DEI laws
Activist Stacey Abrams spoke about the importance of diversity laws and the dangers of removing protections for discriminated groups. She argued that a new ballot question in Wisconsin could stop the government from fixing past mistakes in education and healthcare. Abrams also discussed her book on AI in healthcare and said that technology is a tool that needs careful management. She believes data centers should not be built without community input and that regulations are needed to guide their growth.
Apple removes waitlist for its new Siri AI update
Apple has made its updated Siri artificial intelligence available to most users without a waiting list. The tech giant recently overhauled the voice assistant to improve its capabilities. This change allows customers to access the new features immediately instead of joining a queue. The update represents a major shift in how Apple delivers its AI tools to the public.
UC Berkeley Law starts new course on AI in litigation
UC Berkeley Law School is launching a new course to teach students about using artificial intelligence in legal cases. The class will cover the benefits of AI technology as well as its limitations and risks. Students will learn how to identify the right situations to use these tools in their future legal work. This initiative aims to prepare the next generation of lawyers for an AI-driven legal landscape.
BriteCore hosts webinar on AI in insurance operations
BriteCore is hosting a webinar on November 10 to discuss the future of artificial intelligence in property and casualty insurance. The event will show how AI copilots and agentic workflows can automate complex tasks for insurance companies. CEO Ray Villeneuve explained that the company plans to move from simple assistants to systems that coordinate entire business processes. Attendees will see a live demonstration of how these tools help professionals work faster and manage data better.
Microsegmentation stops AI attacker in Anthropic security test
An Anthropic cyber security evaluation tested whether an AI model could spread through a network after an initial breach. Zero Networks set up a lab with Windows, Linux, macOS and Kubernetes systems and left one machine exposed. The AI model, identified as Claude, attempted 18 attack paths using 23 tools and 22 MITRE ATT&CK techniques. It gained administrator access to the exposed machine but could not move beyond the segmented environment. The test showed microsegmentation contained the attack despite repeated lateral movement attempts.
Why AI cannot truly think like humans do
AI systems such as ChatGPT, Claude, Google Gemini and Perplexity AI can generate text that seems human, but their processes differ from human thinking. Humans use predictive processing based on real world experience, such as anticipating what happens when picking up a cup. AI chatbots learn patterns from large datasets and can describe plausible outcomes without physical experience. Emerging world models may help AI systems anticipate events more like humans, but they still lack true perception, learning and understanding. A shared prediction does not mean AI and humans use the same knowledge or reasoning.
Sources
- Seattle Public Schools tightens student access to YouTube, Minecraft, and AI
- Google’s new AI ‘ultimatum’ sparks concern over tech giant’s power in California schools
- Google turns off AI for schoolchildren by default
- Opinion | We Banned Bots From Our Dating Apps. This Is Why.
- 8 Questions on On-Prem AI Infrastructure with Go.AI’s David Moscatelli
- FINRA's AI rules expose gaps in compliance oversight
- Stacey Abrams on DEI, artificial intelligence and regulation ahead of Democrat midterm wins
- Apple mostly drops the waitlist for its new Siri AI
- UC Berkeley Law launches AI litigation course
- BriteCore to Host Webinar on the Future of AI in P&C Insurance, From Copilots to Agentic AI
- AI attacker blocked by microsegmentation in Anthropic test
- Why AI cannot ‘think’ like humans do
Comments
Please log in to post a comment.