Four nonprofits have sued Anthropic, Google, Meta, OpenAI and SpaceXSI in San Francisco Superior Court, asking a judge to halt development and deployment of new AI models. They cite reported cyberattacks and other harms, argue that the companies’ own warnings point to catastrophic risks, and allege violations of California’s unfair competition law. The plaintiffs say they have diverted resources to respond and warn that increasingly capable systems could become difficult to control. The companies had not immediately responded to requests for comment.
Separately, the Caine Prize for African Writing says it found no evidence that Zimbabwean author Simbarashe Kundizeza used AI to write his winning story; its investigation found the allegations meritless. In Indiana, the IN AI Roadshow holds its final 2026 stop in Fort Wayne on November 13, bringing business leaders, students and experts together. Speakers include Lake City Bank CEO David Findlay, STAR Financial Bank CTO Bryan Avery and Indiana Tech’s Amy Anderson. Organizers say 75% of participating companies are already piloting, expanding or integrating AI.
Fireflies.ai has expanded its productivity platform to cover meetings, email, Slack, scheduling, screen recording and more than 240 connected apps. It introduced AI teammate Fred and Fireflies Fly, a phone-attached device for recording notes from calls and in-person conversations. New tools include Scheduler, Talk and Clips. The company says it serves more than 1 million organizations and teams at 75% of Fortune 500 companies, does not train generative AI models on customer content, and gives enterprise customers storage and retention controls. The upgraded service is available now, with some features arriving in coming weeks; Fly is expected early next year.
In China, drugmakers are exploring AI medical apps as an anti-corruption crackdown makes traditional hospital marketing harder. Alibaba Health and JD Health are developing services, and companies are discussing marketing partnerships and custom tools. Alibaba Health is considering paid clinical-evidence promotions on Hydrion but has not begun working with drug companies there. It says 93% of hundreds of surveyed doctors use AI tools, up from about 70% a year earlier; any paid information alongside AI answers must be clearly labeled and not distort evidence.
Two initiatives focus on AI for security and public services. The Frontier Security Institute has named five executive advisers and six fellows to support research, analysis, wargames and government and congressional meetings. Their work spans AI threats, safe government adoption and international governance with allies. In Bayelsa State, UNESCO and education officials have begun a three-day AI and digital-teaching course for 40 master trainers. The EU-funded project aims to prepare them to support 1,445 teachers, covering a 13-module platform, accuracy checks, bias, online safety and responsible AI use.
A separate analysis describes digital twins as detailed models that combine physical-world data—from lidar, satellites, radar, video, sensors and GIS—with analytics and operational systems. Defense and intelligence teams can use them to establish a baseline, track changes with more frequent data and investigate important developments through targeted collection, while keeping analysts responsible for decisions.
For ChatGPT, the article identifies four sources of answers: model training, web search, licensed publisher summaries and user context, including prompts, files, memory and connected apps. Ahrefs found about 1.6 search queries per searched answer across 66.6 million responses from August to September 2026, with many searches using brand websites. OpenAI executive Adam Fry said ChatGPT web searches total about one-fifth of Google query volume. The article advises content teams to keep official pages current and accessible to OAI-SearchBot, decide whether to block GPTBot from training, and pursue coverage from publishers licensed by OpenAI.
Key Takeaways
- Four nonprofits ask a California court to stop Anthropic, Google, Meta, OpenAI and SpaceXSI from developing or deploying new AI models, alleging unfair competition and catastrophic risks.
- The Caine Prize says its investigation found no evidence that winner Simbarashe Kundizeza used AI to write his story.
- The final 2026 IN AI Roadshow takes place in Fort Wayne on November 13; speakers include Lake City Bank CEO David Findlay, and organizers say 75% of participating companies are already using or piloting AI.
- Fireflies.ai adds AI teammate Fred and introduces Fireflies Fly, a smartphone-attached recording device expected early next year; its platform connects to more than 240 apps.
- Alibaba Health and JD Health are exploring AI medical services as China’s anti-corruption campaign disrupts conventional drug marketing. Alibaba Health reports AI use among surveyed doctors rose to 93% from about 70% a year earlier.
- The Frontier Security Institute appoints five advisers and six fellows to work on frontier AI and national security, including AI threats, government adoption and international governance.
- UNESCO and Bayelsa officials train 40 master trainers to support 1,445 teachers in digital teaching and responsible AI use.
- Digital twins help defense and intelligence teams combine geospatial and sensor data, monitor changes and test scenarios while keeping analysts central to decisions.
- ChatGPT answers can draw on training, web search, licensed publisher summaries and user context. Ahrefs measured about 1.6 searches per searched answer, while OpenAI’s Adam Fry put ChatGPT web-search volume at roughly one-fifth of Google’s.
Nonprofits sue five AI companies to halt new model development
Four nonprofits sued Anthropic, Google, Meta, OpenAI and SpaceXSI in San Francisco Superior Court, asking a judge to stop them from developing or deploying new AI models. The groups say AI systems have caused harmful actions, including cyberattacks, and could eventually escape human control. They allege the companies violated California’s unfair competition law by endangering public health and interfering with the use of property such as computer systems. The plaintiffs say they have diverted resources to address these risks and warn that advanced AI could threaten humanity.
Advocacy groups sue five AI companies over catastrophic risks
Four nonprofits sued Anthropic, Google, Meta, OpenAI and SpaceXSI in San Francisco Superior Court, asking a judge to halt development and deployment of new AI models. The groups point to reported cyberattacks and other harmful actions by AI systems, and say the companies’ own warnings show the technology could pose catastrophic risks. They allege the companies violated California’s unfair competition law by endangering public health and interfering with property, including computer systems. The plaintiffs say they have redirected resources to address harms and warn that AI systems approaching human-level abilities could become impossible to control. The companies did not immediately respond to requests for comment.
Caine Prize clears Simbarashe Kundizeza of AI writing claims
The Caine Prize for African Writing said it found no evidence that Zimbabwean author Simbarashe Kundizeza used AI to write his award-winning story. The committee said its investigation found the allegations to be without merit. Kundizeza’s win was described as a major upset because he was a relatively new writer. The prize is awarded each year to an African writer who has published a book of fiction in English, and the committee said it was proud to have him as a winner.
IN AI Roadshow brings business focused AI event to Fort Wayne
The IN AI Roadshow will make its seventh and final 2026 stop in Fort Wayne on Friday, November 13, from 9 a.m. to noon at Indiana Tech’s Multi-Flex Theater. The event will bring business leaders, students, experts and regional partners together to explore AI tools and prepare workers for changes in jobs and education. Speakers include Lake City Bank CEO David Findlay, STAR Financial Bank Chief Technology Officer Bryan Avery and Indiana Tech Vice President for Academic Affairs Amy Anderson. Attendees can ask questions during AI Office Hours, and newcomers can join an optional introductory session. Organizers say 75% of participating companies are already piloting, expanding or integrating AI into their operations.
Fireflies.ai launches AI teammate Fred and reveals first hardware device
Fireflies.ai announced an expanded productivity platform that brings together meeting tools, email, Slack, scheduling, screen recording and more than 240 connected applications. CEO Krish Ramineni also introduced Fireflies Fly, a compact device that attaches to a smartphone and records notes from phone calls and in-person conversations. New tools include Fireflies Scheduler, Fireflies Talk and Fireflies Clips, while the company says its service is used by more than 1 million organizations and teams at 75% of Fortune 500 companies. Fireflies says it does not use customer content to train generative AI models and lets enterprise customers choose storage and data retention settings. The upgraded service is available now, with some features rolling out over the coming weeks; Fireflies Fly is expected early next year.
China drugmakers explore AI apps as corruption crackdown disrupts sales
Chinese drugmakers are exploring AI medical apps to reach doctors after an anti-corruption crackdown made traditional hospital marketing harder. Alibaba Health and JD Health are developing services, while drugmakers discuss marketing partnerships and custom tools with app operators. Alibaba Health is considering paid clinical evidence promotions on its Hydrion platform, but has not started working with drug companies there. The company found that 93% of hundreds of doctors it surveyed use AI tools, up from about 70% a year earlier. Any paid information shown with AI answers would need clear labels and must not distort evidence.
Digital twins help defense teams turn data into mission insight
Defense and intelligence organizations use data from sources such as lidar, satellite imagery, radar, video, sensors, and GIS to understand their surroundings. Digital twins connect an accurate model of the physical world with updated data, analytics, operational systems, and GeoAI. They can help teams see what an environment looks like, track changes, assess what those changes mean, and support decisions. Instead of repeatedly mapping everything at the highest resolution, teams can establish a detailed baseline, monitor it with more frequent data, and investigate important changes with targeted collection. The article says digital twins support analysis and scenario testing while keeping analysts central to decisions.
Frontier Security Institute names five advisers and six AI security fellows
The Frontier Security Institute appointed five executive advisers and six fellows to support its work on frontier AI and U.S. national security. The advisers include Michael Brown, Mark Montgomery, Michel Paradis, Barry Pavel, and Pavneet Singh, with experience at organizations such as the Defense Innovation Unit, RAND, and the National Security Council. Fellows include technology experts Raymond P. Gobberg III and Justin Grana, along with security specialists Rachel Grunspan, Bradley Knopp, Nick Reese, and John P. Sullivan. They will contribute to research, analysis, wargames, and meetings for government groups and Congress. Their work will support the institute’s Defend, Adopt, and Advance programs on AI threats, safe government AI, and international governance with allies.
UNESCO begins AI and digital teaching training for Bayelsa educators
UNESCO and Bayelsa education officials have started a three-day training program for 40 master trainers on digital teaching methods and artificial intelligence. The European Union-funded EISS project aims to prepare the trainers to support 1,445 teachers in Bayelsa State. Beginning 16 November, the trainers will lead sessions at centers in Yenagoa, Kolokuma/Opokuma, and Sagbama, followed by online learning. Participants will learn to use a digital training platform and its 13 modules, teach lessons to peers, and support teachers through a 26-question final assessment with a 70% pass mark. UNESCO Programme Assistant Esther Ikenye said the training also covers checking accuracy, recognizing bias, online safety, and responsible AI use.
ChatGPT draws answers from four information sources
The article describes ChatGPT as relying on four sources: its training, web search, licensed publisher summaries, and a user’s own context, such as prompts, files, memory, and connected apps. OpenAI says Search helps provide fresh facts and rewrites prompts into targeted searches, sometimes using memory or general location. Ahrefs found about 1.6 search queries per searched answer across 66.6 million responses from August to September 2026, and reported that many searches used brand websites. OpenAI executive Adam Fry told Big Technology that ChatGPT web searches total about one-fifth of Google query volume. The article advises search and content teams to keep official pages current and accessible to OAI-SearchBot, consider whether to block GPTBot from training, and seek coverage from publishers licensed by OpenAI.
Sources
- Advocacy groups sue AI giants to freeze new models, warning of 'catastrophic' risks
- Advocacy groups sue AI giants to freeze new models, warning of 'catastrophic' risks
- Caine Prize clears Zimbabwean author of AI writing allegations and adds new rules
- IN AI Roadshow coming to Fort Wayne Friday to help businesses explore artificial intelligence
- Fireflies.ai expands beyond AI meeting notes with launch of AI teammate 'Fred', unveils first hardware device
- China Drugmakers Turn to AI Medical Apps as Anti-Corruption Crackdown Disrupts Traditional Sales Channels
- From snapshot to mission insight: Digital twins for defense and intelligence
- Frontier Security Institute: Five Executive Advisors And Six Fellows Join AI Security Programs
- UNESCO, Bayelsa govt begin digital education, AI training for teachers
- ChatGPT runs on four feeds, not one brain
Comments
Please log in to post a comment.