AI regulation is heating up in Pennsylvania, where Governor Josh Shapiro is calling for swift government action and warning against trusting tech companies to self-regulate. His stance follows public warnings from researchers and leaders at Anthropic, OpenAI, and Google about the risks of increasingly advanced AI systems. Shapiro's opponent, Stacy Garrity, accuses him of shifting his position on the issue as the race intensifies.
AI governance experts are amplifying these concerns, cautioning that unchecked AI development could lead to catastrophic outcomes. They urge policymakers to establish clear guidelines now, balancing innovation with long-term safety considerations for humanity.
Anthropic specifically warned that AI systems are growing capable of building future versions of themselves, raising further alarms about control. A separate analysis asks when AI might stop needing humans entirely, noting the real threshold is not just self-copying but recreating the full infrastructure — energy, hardware, supply chains — that keeps it running.
On the practical side, Dutch telecom KPN partnered with McKinsey's QuantumBlack to deploy an agentic AI system handling millions of customer calls. The system responds in under two seconds, allows interruptions, and includes guardrails with daily reviews of up to 100 call transcripts. Employee adoption surpassed 86%, and human agents are now freed to tackle more complex problems.
Meanwhile, ASUS showcased its AI infrastructure plans at AI Tech 2026 in Seoul, emphasizing that AI demands more than computing power — it needs better energy efficiency, cooling, connectivity, and automation. In other developments, Algerian entrepreneur Nacer Abdelli founded EduGato to use AI for matching job candidates with openings, and a new editorial argues AI's logical capabilities can never replace human emotion and wisdom in decisions about life, justice, or faith.
On the international front, a column advocates for expanded AI safety talks with China, noting that since 2023, Chinese and Western groups have held about a dozen informal discussions. The author calls for more meetings as AI systems grow more capable of improving themselves.
Key Takeaways
- Governor Josh Shapiro pushes for fast AI regulation in Pennsylvania, rejecting tech company self-regulation
- Anthropic, OpenAI, and Google leaders have warned about risks of advanced AI systems
- AI governance experts warn unchecked AI growth could lead to catastrophic consequences
- Anthropic announced AI systems can now build future versions of themselves
- KPN and McKinsey's QuantumBlack built an agentic AI customer service system handling millions of calls with under-two-second response times and over 86% employee adoption
- ASUS showcased AI infrastructure at AI Tech 2026 in Seoul, focusing on energy efficiency, cooling, and automation
- Nacer Abdelli's EduGato platform uses AI to match candidates' skills with job openings in Algeria
- An editorial argues AI cannot replace human emotion and should never be the final authority in life, justice, or faith decisions
- Chinese and Western groups have held about a dozen informal AI safety discussions since 2023
- Experts say the key threshold for AI independence is not self-copying but recreating full infrastructure including energy and supply chains
AI debate heats up in Pennsylvania governor race
Artificial intelligence has become a major topic in the Pennsylvania governor's race. Governor Josh Shapiro is calling for new AI regulations. His opponent, Stacy Garrity, says he is changing his position on the issue. The debate is intensifying as the campaign continues.
Josh Shapiro pushes for fast AI rules and warns against trusting tech companies
Pennsylvania Governor Josh Shapiro is urging quick action on AI regulation. He says relying on tech companies to self-regulate is flat-out wrong. His comments follow warnings from researchers and leaders at Anthropic, OpenAI and Google about the risks of advanced AI systems. Shapiro wants the government to step in and set clear rules.
AI governance experts warn of dangers and call for regulation
AI governance experts are warning that unchecked AI growth could lead to catastrophic consequences. Dr. Rachel Kim said the risks are very real and action is needed now. Dr. John Taylor and Dr. Maria Rodriguez also urge policymakers to think about long-term effects. They want clear guidelines for AI development and use to prevent disaster while still considering the future of humanity.
Dutch telecom KPN builds AI system for millions of customer calls
Dutch telecom company KPN partnered with McKinsey and its AI arm QuantumBlack to build an agentic AI system for customer service. The system handles voice conversations with response times under two seconds and lets customers interrupt agents mid-sentence. KPN introduced guardrails and daily reviews of up to 100 call transcripts to monitor quality and safety. The company says employee adoption reached over 86% and AI has reduced handling times while allowing human agents to focus on complex problems.
Anthropic warns AI systems can build future versions of themselves
AI company Anthropic announced that AI systems are increasingly capable of building future versions of themselves. This statement adds to growing concerns about the dangers of powerful AI technology. The announcement came on Thursday from the San Francisco-based company. It highlights mounting worries about AI advancing beyond human control.
When will AI no longer depend on humans to survive
A new article compares how humans domesticated wheat to how we now depend on AI. It asks when AI might stop needing humans for its survival. The key threshold is not just AI copying itself but recreating the entire system that supports it. This includes energy, hardware repair, supply chains, and manufacturing. The concern is that humans may become unnecessary as AI automates the functions it currently relies on us for.
ASUS shows AI infrastructure plans at Seoul tech event
ASUS held AI Tech 2026 in Seoul to showcase infrastructure for the next era of AI. The event stressed that AI needs more than just computing power. It also needs better power use, cooling, connectivity, and automation. ASUS displayed enterprise AI servers, HGX training systems, and rack-scale AI POD deployments. The company emphasized reducing costs and improving efficiency for AI factories.
AI logic cannot replace human emotion editorial says
An editorial compares AI to Mr. Spock from Star Trek to show its limits. AI can act logical and compassionate but it does not truly feel emotions. It cannot love, fear, forgive, or understand sacrifice. The article warns that AI should help humans make decisions but never become the final authority in matters of life, justice, or faith. It says civilization needs wisdom and compassion, not just intelligence.
Nacer Abdelli uses AI to connect training and jobs
Nacer Abdelli is an Algerian entrepreneur who founded EduGato in 2020. The platform uses artificial intelligence to analyze candidates and match their skills with job openings. It combines professional training and recruitment in one place. The goal is to close the gap between education and employment. Abdelli wants to make job searching easier for both workers and employers.
Column on how to approach AI safety talks with China
This column discusses how to handle AI safety discussions with China. It says the goal is not to hold China back but to build trust through transparency. Track 1.5 and 2 talks allow officials and experts to meet informally. Since 2023, Chinese and Western groups have held about a dozen such discussions. The article calls for more meetings as AI systems become more capable of improving themselves.
Sources
- AI becomes debate point in Pennsylvania as governor race begins to heat up
- Josh Shapiro urges swift AI rules, says relying on tech companies is ‘flat-out wrong’
- AI Governance Experts React to Existential Warnings, Calls for Regulation
- Dutch Telecom Builds an AI System for Millions of Customer Calls
- SAN FRANCISCO, California — Artificial intelligence (AI) giant Anthropic said on Thursday that AI systems were increasingly capable of building future versions of themselves, adding to mounting…
- When Does AI Stop Needing Us?
- ASUS AI Tech 2026 Seoul: Building the Infrastructure for the Next Era of AI
- AI’s logic is no substitute for human emotion (Editorial)
- Nacer Abdelli Uses AI to Link Skills Training With Recruitment
- Column: Practice the art of not making an AI safety deal with China
Comments
Please log in to post a comment.