Latest AI (Artificial Intelligence) News
AI Integration into Nuclear Command Systems Poses Severe Global Security Risks
Integration of AI into nuclear command, control, and communications (NC3) accelerates decision-making to machine speed, increases cyberattack vulnerability, and erodes strategic stability. A UN report highlights unprecedented risks and calls for international collaboration on AGI limits in nuclear systems.
The General Assembly launched an Independent International Scientific Panel on AI and Global Dialogue on AI Governance.
US Military's AI Use in Iran War Sparks Ethical Debates and Tech Pushback
The US military reportedly employs advanced AI for intelligence and battlefield operations in the Iran conflict, raising ethical questions. AI firm Anthropic clashed with the Pentagon over concerns on mass surveillance and autonomous weapons.
Security analysts stress humans must retain control of life-and-death decisions for accountability.
Pentagon Designates Anthropic as Supply Chain Risk Over National Security Fears
The Pentagon labeled AI company Anthropic a supply chain risk, potentially forcing government contractors to stop using Claude amid Iran war tensions. President Trump and Defense Secretary Hegseth accused Anthropic of endangering security after CEO Dario Amodei refused concessions on surveillance and weapons.
A six-month phase-out is ordered, prioritizing warfighter tools during combat.
New AI Foundation Model LFM2-2.6B Speeds Up Drug Discovery Pipeline
Insilico Medicine and Liquid AI unveiled a 2.6B-parameter AI model matching larger systems in property prediction, molecular optimization (98.8% success), and retrosynthesis. Trained on 120 billion pharmaceutical tokens, it runs on private infrastructure for ADMET screening and lead optimization.
The model outperforms rivals on multiple benchmarks, aiming to compress discovery timelines.
AI Large Models Revolutionize Catalyst Discovery for Clean Energy
Tohoku University review details AI with MLIPs and LLMs exploring chemical spaces to predict catalyst performance for fuel cells and hydrogen production. This shifts from trial-and-error to self-improving simulations, accelerating innovation in sustainable tech.
Future closed-loop platforms could integrate prediction, synthesis, and testing.
AI-Designed Diffractive Optical Processors Enable Low-Power Structural Monitoring
UCLA's AI-co-optimized diffractive layers encode structural vibrations into light for energy-efficient health monitoring of buildings. Validated on shake tables with seismic waveforms, it extracts 1D/2D spectra using minimal detectors and neural networks.
Scalable to other wavelengths, it shifts computation to physical domain.
AI Poised to Unlock Cancer Complexities with Shared Data Infrastructure
Cancer Research Institute urges standardized global datasets for AI to analyze multimodal cancer data, accelerating immunotherapy breakthroughs. Fragmented silos hinder progress; harmonized networks could simulate experiments and predict treatment responders.
Collaboration across scientists and clinicians is essential for AI-ready research.
OpenAI Agrees to Enhanced Safeguards After Tumbler Ridge Shooting Incident
Canada's AI minister announced ChatGPT will strengthen safeguards following the Tumbler Ridge mass shooting where the shooter evaded flagging. This addresses failures in platform monitoring for violent content or threats.
The move aims to prevent future oversights in high-risk scenarios.
Canadian Government Releases Top 10 AI Security Actions Primer
The Canadian Centre for Cyber Security issued ITSAP.10.049, outlining top actions to bolster AI cyber resilience for organizations. It provides practical guidance across sectors to mitigate AI-specific threats.
Designed for all sizes, it strengthens overall defenses.
87% of Life Sciences Researchers Now Use AI, Demanding New Oversight
A 2025 report shows AI usage in life sciences rose to 87% from 75% in 2023, transforming research tasks. Oversight structures lag behind rapid adoption in the industry.
Compliance frameworks must evolve to manage risks.
New US Legislation to Database Public Lands' Biological Data for AI/Bio Breakthroughs
Senate bill creates a comprehensive database of undiscovered biological data from US public lands to fuel AI-driven biotech discoveries. Announced March 5, 2026, it supports transformative research.
Aims to unlock next-generation innovations.
SXSW 2026 Highlights AI-Powered Scientific Revolution in Discovery Engine
SXSW session explores AI accelerating research in science, healthcare, and creatives via faster experimentation. Showcases reshaping of discovery processes.
Focuses on dramatic speed gains across fields.