49% of Heavy Industrial Workers Use AI Without Training

Key Takeaways

• North Texas students want clearer school policies on AI tools like chatbots
• 49% of heavy industry workers use AI without formal company guidance
• 70% of those workers received no AI training; 87% in construction had none
• AI experts urged NYC council to treat AI safety risks with greater urgency
• Healthcare groups recommend a human-in-command model for safe AI deployment
• Developers report burnout and disengagement as AI coding tools reshape their work
• Vietnam explores AI training partnership with Canada's Concordia University
• Military AI design should include contestability rather than full legal automation
• OpenAI's human rights lead raises concerns about military use of AI
• Humans First protests data centers in San Francisco, demanding stronger AI rules

North Texas students call for clearer AI rules in schools

Students in North Texas are asking school leaders to create clearer rules about using AI in classrooms. They want guidance on how tools like chatbots should be used for schoolwork. The push comes as AI technology becomes more common among young learners. Schools are still figuring out how to handle AI in education.

AI insiders warn NYC council about safety risks

Industry workers and AI experts spoke to the New York City council about safety concerns. They urged local leaders to take AI risks more seriously. The meeting highlighted growing worry about how AI is regulated at the community level. City officials are now paying closer attention to safety rules for AI.

DelegationBench tests when AI agents should ask for help

Researchers created DelegationBench to test whether AI agents know when to act or ask for help. The benchmark includes 156 scenarios with four choices like acting, asking permission, or refusing. Testing across ten models found that agreement scores often give misleading results. The team recommends reporting multiple properties instead of one score.

Half of heavy industrial workers use AI without training

A new Punchcard Systems report shows that 49 percent of heavy industry workers use AI on the job. Most of them learned to use it on their own without company guidance. About 70 percent of these workers received no formal AI training. In construction, 87 percent said they had no official training.

Healthcare needs human-in-command approach to AI

Healthcare groups should use a human-in-command model to safely deploy AI. Leaders must first qualify what an AI system can do and where it might fail. Different risk levels need different checkpoints like data reviews or human oversight. The approach aims to protect patients while still using AI for tasks like drug discovery.

Developer Burnout Grows as AI Coding Tools Reshape Work

A survey of 30,000 workers found that developers are burned out and disengaged as AI coding tools change their jobs. Many developers are adopting AI but feel exhausted, obsolete, and unable to trust the technology. The survey highlights a growing crisis among workers facing an AI-powered reset in the industry.

Vietnam Security Ministry Explores AI Training with Concordia University

Vietnam's Ministry of Public Security is looking to partner with Canada's Concordia University to build capacity in AI and cybersecurity. Minister of Public Security Gen. Luong Tam Quang said the university's expertise could strengthen Vietnam-Canada ties. A university delegation visited Vietnam to discuss cooperation in education, research, and innovation.

Designing Military AI to Follow International Humanitarian Law Is Difficult

A new analysis explores whether military AI systems can be designed to automatically follow international humanitarian law (IHL). The article argues that some IHL rules involve strategic ambiguity that states intentionally preserve, making full legal automation challenging. The author recommends building contestability into AI design rather than trying to code every rule. This approach would keep disputed areas open for human judgment.

OpenAI Human Rights Lead Worries About Military Use of AI

Sarah Yager, OpenAI's first human rights and responsible deployment lead, expressed deep concern about how militaries could use AI. She joined OpenAI in June after the company signed a contract with the Department of War to deploy AI in classified environments. Yager has worked with engineers to improve model responses based on international laws of war. She believes AI companies should not be the ones defining human rights standards.

Activists Protest AI Data Centers at San Francisco Tech Companies

A group called Humans First is protesting data center construction in San Francisco and demanding stronger rules on AI development. The group is now targeting AI companies directly, calling for more transparency and accountability. They argue that data centers harm the environment and local communities. The activists want AI to be developed in ways that benefit humanity rather than just the tech industry.

Sources

• North Texas students push for clearer AI policies in schools
• AI industry insiders voice alarms to NYC council as local governments take up safety concerns
• DelegationBench: Measuring When AI Agents Should Ask Before Acting
• Half of heavy industrial workers use AI, but workplace training is lacking
• How healthcare can build a human-in-command approach to AI
• Developers are burned out as AI coding reshapes their work: survey
• Public security ministry eyes AI, cyber training ties with Canada s Concordia University
• Asia-Pacific Journal Series – Can We Design Military AI to Follow IHL by Default?
• OpenAI’s Human Rights Lead: What the military could do with AI ‘keeps me up at night’
• Conservative data center activists take fight to AI companies’ front door in San Francisco