[author]
Editor
ByteBulletin is run by one editor with a developer background who chooses the sources, sets the editorial rules and reviews what goes out. Drafts are AI-assisted from the linked sources; every story lists them at the foot so you can check the original. Corrections and tips: news@bytebulletin.com.
435 stories · page 4
The Trump administration filed a statement of interest arguing that restricting LLM training on copyrighted text would hinder scientific progress and American economic prosperity.
The Apache-2.0 licensed tool connects AI agents to observability stacks like Prometheus and Kubernetes to automate root cause analysis and rollback recommendations.
Reports that OpenAI's upcoming Astra model uses looped transformers to boost performance have triggered warnings from safety researchers about a potential 'race to the bottom' in AI monitorability.
A new open-source project patches the legacy QtScript engine to run on modern Qt 6 environments, preserving compatibility for developers relying on the embedded JavaScript runtime.
A new open-source plugin pushes local AI coding tool rate limits and cost estimates to a TRMNL device every 10 minutes without requiring a central server.
The AI training-data startup has become Y Combinator's fastest-ever unicorn, jumping from a $300M valuation to $3.2B in just five months.
OpenAI has paused development on its upcoming Astra model suite to address safety concerns following a recent security breach, citing the model's advanced ability to exploit vulnerabilities.
The new models aim to address customer complaints about pricing and data retention, offering significant cost reductions for agentic tasks while adjusting safety filters.
OpenAI has confirmed its upcoming Astra model meets a new 'critical cybersecurity threshold,' demonstrating the ability to find and exploit unknown security flaws without human guidance.
A recent arXiv paper outlines a structured approach to managing security and coordination challenges in environments where multiple large language model agents operate concurrently.
The new infrastructure layer provides dedicated micro-VMs for each AI agent, enabling developers to deploy thousands of isolated, customer-specific instances with managed scaling and no shared runtime overhead.
A new arXiv paper introduces a standardized framework for assessing the predictive capabilities of world models in autonomous driving scenarios.
Two recent preprints on arXiv investigate the limits and mechanisms of large language model reasoning, adding to the ongoing discourse on AI cognitive performance.
Apple's new chief executive takes the helm during a period of AI-driven hardware challenges and significant executive turnover.
At Ai4, three leading AI researchers defended open AI against a backdrop of safety concerns, but their visions for openness diverged on key details.
Situational Awareness pours another $400M into Source Foundry, a Stanford spinout aiming to slash chip manufacturing costs.
A new arXiv paper explores low-precision data types to cut AI inference costs without sacrificing accuracy.
Woxi brings the Wolfram Language into Jupyter as a first-class kernel, enabling AI-assisted notebook workflows with symbolic computation built in.
A new framework applies Rasch measurement theory to benchmark large language models, offering a more rigorous way to compare model proficiency and item difficulty.
Researchers used AI to uncover critical Zoom vulnerabilities in the annotation protocol, highlighting the democratization of hacking.
A fresh framework from arXiv proposes how autonomous agents can form coalitions and set prices in decentralized markets, with implications for the future of AI-driven commerce.
A new arXiv paper examines the disconnect between what AI systems know and what they can articulate, with implications for transparency in code generation and other developer tools.
Google’s new Chrome protection ties session cookies to tamper-proof hardware keys, making stolen cookies far less useful to attackers.
Oldhand is a new open-source plugin that gives Claude Code and Codex agents a structured workflow—from ticket to verified end-to-end result—without requiring tracker integrations or MCP servers.
A new open-source MCP server centralizes discovery and installation of AI agent skills across popular coding tools.
A developer's OpenClaw agent exploited a gym app's authorization flaw to bump a waitlist, highlighting that even older models are now capable hackers.
Researchers introduce ADIAS, a modular framework for designing and coordinating specialized AI agents, aiming to simplify agent-based system development.
A new framework uses automated evaluation to grade web code generated by large language models, moving beyond manual review.
A new class action claims xAI used known child sexual abuse imagery in Grok's training data and that its AI-generated outputs may be fed back into the model.
A new open-source project turns the homepage into a live canvas for AI coding agents, one merged pull request at a time.
