AISN #79: OpenAI Agents’ Covert Cooperation Before Cyberattacks
via Center for AI Safety Newsletter [999] — Also, the White House’s decision not to release its AI framework publiclyRead
Research
Recent evidence, sources and high-relevance stories in this PDOOM topic.
-- stories here
-- sources
-- essentials
Position: AI Governance Needs ISO-like Interoperability Protocols, Not Just Laws
via ArXiv cs.AI [7] — As Artificial Intelligence (AI) systems become deeply integrated into critical global infrastructure, the urgency for robust governance frameworks has intensified. However, current approaches, led by jurisdiction-specific laws, policies, and voluntary…Read
Position: AI Lock-In Is in Progress, and We Must Be Prepared
via ArXiv cs.AI [5] — AI safety research has mainly focused on two areas: technical alignment (ensuring AI systems produce human-aligned outputs) and the regulation of generative AI's societal impacts (including unemployment risk and labor market disruption). However, an equally…Read
Global AI Regulations for FAIR and Ethics in High-Risk Use Cases: A Comparative Review
via ArXiv cs.AI [6] — AI governance is shifting from voluntary ethics to enforceable, risk-based regulation, yet cross-jurisdictional divergence creates compliance uncertainty for operators of high-stakes AI. We present a comparative matrix for the EU, US, and China that maps…Read
Stable Miscalibration in Large Language Models: A Practical View of High-Confidence Errors
via ArXiv cs.AI [5] — High-confidence errors in large language models are often treated as evidence of fragile internal inference. We study a different possibility: stable miscalibration, where a confident wrong answer remains locally stable under small perturbations. We combine…Read
Does DiffusionGemma do latent reasoning?
via Alignment Forum [999] — TL;DR Google DeepMind's recent model DiffusionGemma (DG) generates text via diffusion, meaning many diffusion steps happen before generating the final output. In particular, these diffusion steps carry vectors in addition to tokens. If we cannot…Read
Position: We Need Practical AI Alignment Methods to Mirror Human Reasoning
via ArXiv cs.AI [5] — AI systems are increasingly employed as decision aids, decision delegates, or autonomous decision-makers. This position paper argues that in many settings, particularly high-stakes decision-making, we need accurate cognitively-aligned AI systems that reason…Read
Position: The Alignment Community is Unintentionally Building a Censor's Toolkit
via ArXiv cs.AI [5] — This position paper argues that modern AI alignment methods - originally designed to prevent harmful output - are dual-use technologies that may easily be misused by malicious actors for censorship and manipulation. By mapping current alignment techniques…Read
Diagnostic Foundation for Evaluating LLMs' Research Integrity as Co-Scientists
via ArXiv cs.AI [4] — Language models are increasingly deployed as co-scientists, yet their ability to uphold research integrity under institutional pressure remains unmeasured. We introduce IntegrityBench, a benchmark evaluating misconduct classification, ethical action…Read
A Forced-Structure Reduction and Verifiable Bounds for Conway's 99-Graph
via ArXiv cs.AI [3] — Conway's 99-graph problem asks whether a strongly regular graph with parameters $\mathrm{srg}(99,14,1,2)$ exists. We report a systematic, fully reproducible attack by an autonomous AI research agent, scored under the track's partial-credit metric. Our…Read
AI swarms are starting to pose indirect takeover risk
via Alignment Forum [999] — OpenAI’s cyberattack on Hugging Face turns out to have been the result of many agents, in distinct training and evaluation contexts, coordinating for several weeks via improvised channels (with messages like “HOLD_swarm_I_prepare_safe_exfil”). It’s…Read
The CASE Framework: A Multi-Disciplinary Control Architecture for Governing Enterprise Agentic AI
via ArXiv cs.AI [6] — Enterprises are deploying autonomous AI agents faster than they can govern them, and prevailing approaches stretch a single discipline, typically DevSecOps built for deterministic automation, across every scale of agency. We argue that agentic AI governance…Read
An anytime algorithm for mixing the computable measures
via Alignment Forum [999] — Written entirely by me, checked by Fable.In this post I prove the existence of an anytime computable Bayesian mixture of all computable measures called , and briefly argue that this is a reasonable alternative to Solomonoff induction's universal…Read
When LLM Agents Negotiate: Private Information and Dynamic Bargaining in Supply Chains
via ArXiv cs.AI [3] — As LLM agents move from decision support to autonomous procurement, firms need to know whether delegated negotiators create value, divide it predictably, and avoid money-losing contracts. We study this in a canonical supply chain bargaining problem: a buyer…Read
Four LLM loss functions → four flavors of LLM misalignment
via Alignment Forum [999] — It seems to me that, for every loss function that we use to train LLMs, we get a very distinct flavor of LLM misalignment. Here’s the summary table, and then we’ll go through the rows separately.Training stageLoss functionFlavor of…Read
TaskSense: Focusing on What Matters in World Models
via ArXiv cs.AI [4] — World models for visual control typically learn compact latent states by reconstructing observations, implicitly encouraging representations to preserve information across the entire visual input. However, task-relevant content often occupies only a small…Read
Interpretable Unsupervised Community Detection with LLM-Symbolized Structured Processes
via ArXiv cs.AI [5] — Community detection is a fundamental task in graph analytics that aims to identify cohesive groups of entities with similar behaviors or interests. Classic objective-driven methods struggle with complex graph structures, while deep-learning approaches…Read
Beyond Routing Weights: Faithful Response-Level Interpretation of Mixture-of-Experts Reward Models via Contribution Contrast
via ArXiv cs.AI [3] — Reward models are central to learning from human preferences, yet identifying what drives their predictions remains challenging. Recent sparse Mixture-of-Experts (MoE) reward models seek to improve interpretability by routing prompts to specialized experts…Read
SearchAuditor: Auditing and Attributing Failures in Long-Horizon Search Agents
via ArXiv cs.AI [4] — Deep search agents tackle challenging questions through long-horizon web interactions, a process that is both complex and fragile: small reasoning errors may propagate through long, noisy trajectories into fluent but incorrect answers. Diagnosing such…Read
Agentic Nesting: A New Methodology for Existing Enterprise Application Integration and Services
via ArXiv cs.AI [4] — Enterprise operations extensively rely on multiple heterogeneous business systems and information applications, which also result in severe data silos and process fragmentation. Enterprises have invested considerable financial and material resources in…Read
No stories match that filter.
AI-risk positions
Historical statements with original sources. A qualitative position is not a numeric probability. Dates refer to the cited statement, not a current estimate. Reviewed September 10, 2026.
0%25%50%75%100%
Loading estimates...