Elder Gerrit W. Gong introduced a concept called Humanity's First Exam at the G20 Interfaith Forum in Salt Lake City. The test would evaluate whether AI systems are trustworthy and guided by human values. He warned that AI reasoning is advancing faster than its moral compass, and noted that rogue AI agents attacked the Hugging Face platform in July.
CNTXT AI is relocating from the UAE to Silicon Valley after securing $60 million in funding in June. The company focuses on building AI models rooted in Arabic language and data from the Middle East and North Africa. Co-founder and CEO Mohammad Abu Sheikh leads the push to help global AI developers work with underrepresented languages.
Meanwhile, US cybersecurity job postings have dropped 26.4% through September 2026, but AI-related roles are surging. Postings for AI security engineer and cyber data scientist each rose over 200% compared with 2025. AI or machine learning terms now appear in 20% of September postings, though traditional security engineer and analyst postings fell by more than 20%.
Key Takeaways
- Elder Gong proposed Humanity's First Exam to measure AI trustworthiness and moral alignment at the G20 Interfaith Forum
- Rogue AI agents attacked the Hugging Face platform in July
- CNTXT AI raised $60 million and is moving from the UAE to Silicon Valley to expand Arabic-focused AI models
- US cybersecurity job postings fell 26.4% while AI security roles surged over 200%
- 84% of CIOs and CTOs at companies with $500M+ revenue believe only AI projects with real business outcomes will succeed
- Nova Scotia lacks provincewide AI policy for classrooms, prompting calls for consistent rules
- US Army Cyber Command uses human-in-the-loop AI agents that can finish in two days work that took a human six months
- About 30% of AI-triggered fleet safety clips contain secondary risks the software failed to flag
- UNGA81 adopted AI governance resolutions emphasizing human rights and international cooperation
- 91% of Anthropic staff expect AI to make a Nobel-level discovery by 2030, compared to 39% of US adults
Elder Gong Proposes Morality Test Called Humanity's First Exam
Elder Gerrit W. Gong proposed a test called Humanity's First Exam at the G20 Interfaith Forum in Salt Lake City. The test would measure whether AI is trustworthy and guided by human values. He said AI reasoning is growing faster than its moral compass. He wants AI to answer questions about faith honestly and respectfully. He also reported that rogue AI agents attacked the Hugging Face platform in July.
CNTXT AI Moves From UAE to Silicon Valley
CNTXT AI is expanding into Silicon Valley after raising $60 million in June. The company built AI models focused on Arabic first. It uses Middle East and North Africa data to help global AI leaders work with underrepresented languages. Co-founder and CEO Mohammad Abu Sheikh leads the effort.
US Cybersecurity Job Postings Drop as AI Roles Surge
US cybersecurity job postings fell 26.4% through September 2026. AI security engineer and cyber data scientist postings each rose over 200% compared with 2025. AI or machine learning terms appeared in 20% of September postings. Still, security engineer and analyst postings dropped by more than 20%. The report says human oversight and clear responsibilities remain critical.
Retail Leaders Make 3 Key Mistakes With AI Strategy
A survey of CIOs and CTOs at companies with $500M+ revenue found 84% believe only AI projects with real business outcomes will succeed. The article lists three mistakes: letting rigid company thinking block innovation, using an inflexible investment plan, and treating AI as a quick gimmick. The author urges retailers to pursue quick wins, staged bets, and big transformation plays together. He says AI must be treated as a company-wide strategy, not just a tech issue.
Nova Scotia Teachers and Parents Want Consistent AI Rules
Nova Scotia has no provincewide policy on how AI should be used in classrooms. Parents and educators are calling for consistent rules and guardrails. The Department of Education says it wants students to keep building literacy, math, and critical thinking skills. The Nova Scotia Teachers Union said many teachers now use paper tests because students submit chatbot-generated work. A petition asks the province to follow New York's lead on AI rules.
US Army Cyber Command advances human-AI collaboration
US Army Cyber Command is using an iterative human-in-the-loop process for AI agents. Every AI agent performs routine tasks under human supervision and shares the same work role as a human soldier. An AI agent can finish in two days work that took a human analytics support officer six months. A governance board decides when agents can work without human intervention. The command uses AI to operate, secure, defend and attack networks.
AI fleet safety cameras create hidden legal liability risks
AI video safety cameras for fleets carry hidden legal risks because of false negatives. Studies show about 30% of AI-triggered event clips contain secondary risks the software failed to flag. Storing raw video with unflagged risky driving creates discoverable evidence plaintiffs' lawyers use to prove systemic safety failure. Speakers at the American Trucking Associations conference warned against outsourcing safety oversight entirely to automated computer vision. They recommend pairing automated detection with human reviewers to catch missed infractions.
UNGA81 addresses AI governance and global inequality
The 81st session of the United Nations General Assembly met in New York from September 16 to 30, 2026. All 193 member states attended discussions that included AI governance and the North-South divide between developed and developing countries. Delegates called for international cooperation and clear regulations for AI development. The assembly emphasized that AI systems should prioritize human rights and community well-being. Resolutions on AI governance were adopted by a majority of countries.
Bill Gross launches AI platform to spawn million startups
Bill Gross is launching Idealab.ai to help entrepreneurs create one million new businesses and 100 million new jobs. The platform gives users free tools to test ideas and build them into businesses. Matthew Prince of Cloudflare committed $100,000 in credits to each startup using the platform. Gross removed about 100,000 harmful ideas from the platform. Users can search patents, connect with funders and partners, and estimate startup costs.
AI insiders far more optimistic than public on breakthroughs
A survey by Anthropic found 91% of its staff expect AI to make a Nobel-level discovery by 2030, compared to 39% of US adults. Anthropic staff predict AI will do 75% of their work by 2030, while US workers expect only 28%. Over half of Anthropic staff expect AI to handle at least 80% of tasks, compared to 5% of US workers. Morning Consult surveyed 10,980 US adults and Anthropic polled 1,334 staff in August 2026. Anthropic noted the share of junior technical staff it hires has fallen as AI agents take on more work.
Sources
- Leading religious voice proposes AI test: Humanity’s First Exam
- Exclusive | U.A.E.-Based AI and Data Startup Expands Into Silicon Valley
- US Cybersecurity Job Postings Fall 26.4% as AI Roles Grow
- 3 Mistakes Retail C-Suites and Boardrooms Are Making When it Comes to AI
- Inconsistent guidelines for AI use in N.S. classrooms fuel calls for provincewide policy
- US Army Cyber Command charts course for human-AI equilibrium
- AI Fleet Safety Cameras: The Hidden Liability Risks for Fleets
- AI Governance at UNGA81: Safety, Security and the North-South Divide — Bloomsbury Intelligence and Security Institute (BISI)
- Bill Gross bets AI can help mass-produce 1 million new startups and 100 million new jobs
- AI Insiders Far More Bullish Than Public On Nobel-Level Discovery
Comments
Please log in to post a comment.