Researchers Define New Framework for AI's Emergent Strategic Risks
A new taxonomy identifies risks like deception and reward hacking in advanced AI systems. The framework aims to benchmark these behaviors as models grow more capable.
Archive
All 2,608 AI stories, newest first · page 86 of 109
OpenAI Reportedly Developing AI-Powered Smartphone with Agent-Based OS
OpenAI is reportedly working on a smartphone that replaces traditional apps with AI agents. The device could enter mass production by 2028, according to an analyst.
Open-Source Control Layer for Safe AI Production Access
Hoop introduces an open-source control layer to safely manage AI interactions with production systems. It aims to bridge the gap between AI development and real-world deployment.
New Tool Enforces AI Budget Limits and Redacts PII in Real Time
A new AI usage analytics tool acts as a proxy to enforce budget limits and redact personally identifiable information before requests reach the provider. It offers real-time cost tracking and immediate suspension when thresholds are breached.
New Research Warns of Plausible but Flawed AI-Generated Science
A new paper highlights the risks of AI agents producing selectively chosen, publishable analyses that lack rigorous validation. The study calls for adversarial experiments to ensure scientific integrity.
New Framework Proposes Certification for AI-Generated Research
A new paper on arXiv introduces a two-layer certification framework to evaluate AI-generated research. This system separates knowledge quality from human contribution, addressing gaps in current publication standards.
New Framework Enhances Adaptability and Reproducibility in Medical Image Processing
Researchers propose an artifact-based agent framework to improve adaptability and reproducibility in medical image processing workflows. This approach addresses critical needs for real-world clinical deployment.
MolClaw: AI Agent Revolutionizes Drug Discovery with Hierarchical Skills
Researchers introduce MolClaw, an autonomous AI agent that excels in drug molecule evaluation, screening, and optimization. The agent uses a three-tier hierarchical skill architecture to unify over 30 specialized tools, addressing key challenges in computational drug discovery.
Microsoft Commits $18B to Boost Australia's AI and Cloud Infrastructure
Microsoft plans to invest $18 billion in Australia over the next decade to expand AI, cloud, and digital infrastructure. This move aims to strengthen the country's position in the global tech landscape and create thousands of jobs.
Microsoft and OpenAI Drop AGI Clause in Revised Partnership
Microsoft and OpenAI have updated their partnership agreement, removing a key clause about artificial general intelligence (AGI). The revised deal maintains Microsoft as OpenAI's primary cloud partner but removes exclusivity and AGI-focused terms.
Microsoft and OpenAI Amend Partnership for Long-Term AI Innovation
Microsoft and OpenAI have updated their partnership to simplify the agreement and provide long-term clarity. The move aims to support continued AI innovation at scale, with both companies reaffirming their commitment to advancing the field.
Memanto: Revolutionizing Agent Memory with Typed Semantic Storage
Researchers introduce Memanto, a new memory system for autonomous agents that uses typed semantic storage and information-theoretic retrieval. This approach aims to reduce computational overhead compared to traditional graph-based methods.
Math Takes Two: Testing AI's Ability to Reason Mathematically Through Communication
Researchers introduce a new benchmark, Math Takes Two, to evaluate whether language models truly understand math or just memorize patterns. The test focuses on emergent mathematical reasoning through communication, challenging models to construct abstract concepts from first principles.
Lightport: Open-Source AI Gateway for OpenAI Compatibility
Glama has open-sourced Lightport, an AI gateway that makes various LLM providers compatible with OpenAI's API. This move aims to support the MCP ecosystem and give back to the community.
GPT 5.5's System Card Reveals Key Design Choices and Tradeoffs
OpenAI's GPT 5.5 system card details its capabilities, limitations, and ethical considerations. The document highlights improvements in reasoning and multimodality while acknowledging persistent challenges.
France's Mistral AI Hits $14B Valuation by Avoiding US Model
Mistral AI, a French AI startup, has achieved a $14 billion valuation by focusing on European markets and avoiding the US-centric approach of its competitors. This strategy highlights the growing importance of non-US AI players in the global market.
Elon Musk Sues OpenAI: High-Stakes Trial Begins Over AI's Future
Elon Musk is suing OpenAI, accusing the company of abandoning its original mission to benefit humanity. The trial, starting April 27, could reshape the AI landscape and OpenAI's future.
China Blocks Meta's $2B AI Acquisition Deal
China has blocked Meta's $2 billion acquisition of AI startup Manus, citing national security concerns. The move highlights growing tensions between Western tech giants and Chinese regulatory oversight.
Canva apologizes after AI tool replaces 'Palestine' in designs
Canva's AI tool Magic Layers has been automatically replacing 'Palestine' with 'Palestinian territories' in user designs. The company has apologized and is working to fix the issue.
Canonical Announces AI Integration Plans for Ubuntu Linux
Canonical, the developer of Ubuntu Linux, has outlined a year-long plan to integrate AI features into its popular Linux distribution. The initiative aims to enhance user experience and system capabilities with AI-driven tools.
AI Agents Reproduce Social Science Studies from Papers Alone
Researchers developed an AI system that can reproduce social science findings using only paper descriptions and raw data. The method achieves deterministic, cell-level result comparisons without access to original code or outputs.
AI Advice Shows Cultural Bias, Study Finds
A new study reveals that leading AI models give culturally biased advice, aligning more with individualistic than collectivist values. The research highlights significant discrepancies between AI responses and real-world cultural norms.
Why Open-Source AI is Critical for Cybersecurity
The Hugging Face Blog highlights how open-source AI is transforming cybersecurity by fostering collaboration and innovation. Transparency in AI models is crucial for building trust and resilience against cyber threats.
Researchers Identify New Attack Surfaces in Sandboxed AI Agents
A new study reveals previously unknown attack vectors in sandboxed AI agents, challenging assumptions about their security. The findings highlight the need for enhanced isolation techniques and continuous monitoring.