Latest AI (Artificial Intelligence) News
2026 Pro-Human AI Declaration Calls for Human Control Over AI
A coalition of AI ethicists and leaders, including Yoshua Bengio and Sir Richard Branson, launched the Pro-Human AI Declaration demanding safeguards like human off-switches, pauses on superintelligence development, and bans on rogue self-replicating systems. 72% of voters support holding companies legally responsible for AI harms.
The declaration addresses fears of AI hallucinations, Singularity, and human replacement.
Anthropic and Pentagon Resume Talks on High-Stakes AI Defense Deal
Anthropic is negotiating a major AI defense contract with the Pentagon amid investor pressure to de-escalate clashes over safeguards. OpenAI's Sam Altman responded to criticism over its own Pentagon AI deal and is eyeing a NATO contract.
These developments highlight intensifying military AI integrations.
OpenAI Annualized Revenue Hits $25 Billion Amid Global Surge
OpenAI achieved a $25 billion annualized revenue milestone driven by widespread global adoption. The company also agreed to new safeguards after failing to flag a shooter to police in Canada.
Microsoft plans to license AI agents like employees.
UK Launches £40M Frontier AI Research Lab for Tech Independence
The UK government announced a £40 million investment in a new frontier AI lab to accelerate breakthroughs and bolster technological sovereignty. This initiative aims to position the UK as a leader in advanced AI research.
Scientists Build Tiny AI Brain Model Using Monkey Neuron Data
Researchers developed a miniature AI brain model trained on monkey neuron data, advancing neuromorphic computing. This innovation bridges biological and artificial intelligence systems.
AI Accelerates Catalyst Discovery for Clean Energy Technologies
Tohoku University researchers outlined an AI platform using MLIPs and LLMs to predict and design catalysts for fuel cells, hydrogen production, and pollution control, drastically reducing discovery time. The system enables self-improving cycles of simulation and experimentation.
Hao Li envisions closed-loop AI platforms expanding to batteries and hydrogen storage.
New AI Foundation Model Speeds Up Drug Discovery on Private Infrastructure
Insilico Medicine and Liquid AI released LFM2-2.6B-MMAI, a 2.6B-parameter model outperforming larger systems in ADMET prediction, molecular optimization (98.8% success), and retrosynthesis. Trained on 120B pharmaceutical tokens, it runs on private setups without cloud data sharing.
CEOs hail it for compressing discovery timelines.
AI-Designed Diffractive Optical Processors Enable Low-Power Structural Monitoring
UCLA researchers created AI-co-optimized diffractive layers for passive, energy-free vibration monitoring of structures like buildings during earthquakes. Validated on shake tables, the scalable system encodes oscillations into light for neural network decoding.
It supports multipoint, wavelength-multiplexed sensing.
World Economic Forum: AI Needs Shared Data Infrastructure to Unlock Cancer Breakthroughs
Cancer Research Institute CEO Alicia Zhou urges standardized global data networks for AI to analyze immunotherapy data, predict responses, and simulate treatments. Fragmented silos hinder progress despite AI's pattern recognition power.
Harmonized datasets could compress decades of discovery into years.
AI Discovered Drugs Accelerate Clinical Timelines in 2026
AI-designed drugs like Insilico's INS018055 are shortening clinical development timelines, shifting AI from productivity to biological impact. This trend underscores AI's growing role in pharmaceuticals beyond traditional methods.
TikTok Rejects End-to-End Encryption Citing AI-Enhanced User Safety
TikTok declined end-to-end encryption to maintain AI-driven safety monitoring features. The decision balances privacy with proactive threat detection.