Kazakhstan Launches AlemLLM and Alem.cloud Supercomputer

Key Takeaways

• <ul><li>Kazakhstan launched AlemLLM and Alem.cloud supercomputer while joining both U.S.-led Pax Silica and China's WAICO for balanced AI governance.</li><li>President Tokayev wants the Washington Model to become the basis for a future UN treaty on AI security.</li><li>Uzbek President Mirziyoyev flagged AI deepfakes, fraud, and extremist propaganda as major security threats in Central Asia.</li><li>Uzbekistan proposed hosting a regional office of the UN Office of Counter-Terrorism to address AI-driven threats.</li><li>Google.org and Clalit Health Services are building an AI-powered Digital Front Door using Gemini AI for 5 million patients in Israel.</li><li>AMA co-founder says each cancer patient produces about 20,000 gene-level data points that AI agents could help analyze, though no AI-designed drug has won FDA approval yet.</li><li>AI skill requirements in jobs have tripled since 2022, but over 60% of employers gave no raise or title change for added AI duties.</li><li>About 19% of U.S. workers replaced by AI were later asked to return, with over half wanting higher pay.</li><li>A military AI tool helped stop a FIFA World Cup bomb threat by completing reports in 90 seconds versus the usual 20 minutes.</li><li>CubCloud AI sued Missoula over a data center ban, citing Montana's Right to Compute Act.</li></ul>

Kazakhstan balances AI ties with US and China for security role

Kazakhstan is using its partnerships with both the United States and China to push its own AI security agenda. On November 6, Astana hosted the Aqaba Process meeting with Central Asian leaders and King Abdullah II of Jordan. The country joined both the U.S.-led Pax Silica and China's WAICO, splitting roles between the two powers. Kazakhstan launched the AlemLLM model, the Alem.cloud supercomputer, and passed the region's first AI law. President Tokayev wants the Washington Model to become the basis for a future UN treaty.

AI deepfakes and fraud pose new security threats in Central Asia

Deepfakes, online fraud, and extremist propaganda are emerging as major security threats in Central Asia. AI tools now make it easy and cheap to fake voices, videos, and documents, which can spread disinformation quickly. On November 6, Astana hosted a regional meeting on AI security where Uzbek President Mirziyoyev discussed the dangers. Countries in the region are responding with new laws, cybercrime prevention, and public warnings. Uzbekistan has proposed hosting a regional office of the UN Office of Counter-Terrorism.

YC founder says cancer treatment is now a data challenge

Alfredo, co-founder of AMA, says cancer is now a data problem that AI can help solve. He noted that gene sequencing has become very cheap and each patient now produces about 20,000 gene-level data points that hospitals cannot fully analyze. AMA wants to use AI agents to replace large genomics teams and guide treatment decisions. He described cases where AI analysis found better therapy options for patients. However, no AI-designed drug has yet won FDA approval, and some industry leaders remain cautious.

Google and Clalit build AI health assistant for millions of patients

Google.org and Clalit Health Services in Israel announced a two-year grant to build an AI-powered Digital Front Door for 5 million patients. The system uses Google Cloud and Gemini AI to help patients understand their healthcare needs, book appointments, and manage care in their native languages. It uses voice-based tools so patients without internet or smartphones can still access help. Clalit plans to make the voice AI architecture open source so other healthcare systems around the world can use it.

Hackers use AI to power more dangerous cyberattacks on companies

Cybersecurity experts say hackers are using AI to make attacks faster, larger, and more autonomous, putting companies under growing pressure. Legal and compliance teams are advised to brief senior leadership, reassess cyber provider retainers, and review vendor contracts carefully. Experts also warn that organizations should double-check facts before sending privacy notifications during incidents, since some threat actor evidence may be AI-generated. Companies should also watch for public reporting obligations, such as SEC Form 8-K filings within four business days of a material breach.

AI changes work but job descriptions and pay lag behind

AI skill requirements in jobs have tripled since 2022, but most employers have not updated titles or pay. Over 60% of employers adding AI duties gave no raise or title change. Sixty-five percent of HR teams say AI requirements were added to nontechnical roles. Employees often get more work instead of rewards, so companies need to revise roles and pay.

Missoula AI company sues city over data center ban

CubCloud AI sued Missoula over a zoning change that blocks data centers in the city. The company says the ban breaks Montana's Right to Compute Act. CubCloud has 18 employees and runs a facility near the Missoula airport. It wants the court to stop the city from enforcing the new rules.

Workers replaced by AI are being asked back with higher pay demands

About 19% of U.S. workers replaced by AI were later asked to return, a SHL survey found. Over half of those returnees want higher pay. Many also want job security guarantees and clarity on how AI is used in their roles. Companies like Klarna and Ford are rehiring some workers for tasks AI cannot handle well.

The climate case for AI has not been made

AI companies claim the technology can help fight climate change, but progress is slow. A 2024 study projected big emissions cuts from AI in energy, food and transport, but most efforts are still plans. Only one grid operator reviewed has a working AI program. Oil and gas interest in AI and rising data center energy use also make the problem harder.

4 areas where companies need AI guardrails

Companies need AI guardrails in development, deployment, monitoring and auditing. These rules help make AI systems clear, fair and accountable. They prevent wasted money, failed projects and hidden risks. Guardrails in each area help companies get AI benefits while avoiding harm.

Military AI tool stopped World Cup bomb threat

A military-made artificial intelligence tool helped law enforcement stop a bomb threat at the FIFA World Cup in California this summer. The incident happened on June 18 when Switzerland played Bosnia and Herzegovina at SoFi Stadium outside Los Angeles. Army Captain Ryan Hetrick, who leads the Pentagon's Generative AI Task Force, said the AI agent could complete spot reports in 90 seconds compared to the usual 20 minutes. This extra time allowed authorities to coordinate a response and deter the bomb threat. Hetrick has become a military evangelist for AI integration across operations.

Sources

• Kazakhstan Turns Its Dual AI Partnerships Into Regional Security Agenda
• Artificial Intelligence: A New Security Threat to Central Asia?
• YC Founder Says Cancer Is Now a Data Problem
• Google and Clalit to build AI health assistant for 5 million patients
• Thought for the week: Companies face cyber pain as hackers press AI advantage
• AI is rewiring work. Job descriptions and compensation haven’t caught up.
• Missoula AI company sues city over data center restrictions
• ‘AI rebounding’: Nearly 20% of U.S. workers replaced by AI are being asked back—now they’re demanding higher pay, job security and AI transparency
• The climate case for AI has not been made
• 4 areas where companies need AI guardrails
• Military-made AI tool helped thwart World Cup bomb threat