Discourse monitor
How the monitored publications were assessed
Discourse valence across every scored item: what the sampled sources published and how independent models read it. It is not a measure of public opinion, not a measure of the state of the world, and article volume is not agreement.
Mean discourse valence per day · evaluation · last 90 days
Discourse valence, window mean
-30
Adverse
120 items · range -82 to 0
-100
-50
0
+50
+100
-30
Adverse readingFavourable reading
120 items scored in this window
- When Routing Reveals Membership: Privacy Leakage from MoE Router Telemetry
- Move Fast and Mend Things: Keeping Up with Evolving AI Harms Using Social Media Commentary
- Teachers at Euan Blair firm report ‘horrendous stress’ after AI used to rate their work
- I Used AI to Fake a Scientific Paper. Here’s What Happened
- Court Tosses Sentence After A.I. Video of Victim ‘Forgiving’ His Killer Is Played
- Whose Data, Whose Benchmark? A User-Centric Framework for Auditing Evaluation Resources
- Securing the agentic enterprise with data lineage
- Representation Alignment from a Few Paired Data via Distilled Optimal Transport
- Argus: Academic Integrity in the Era of Generative AI
- Talking with Language Models
- Operationalizing NCII and CSAM Law in the Age of AI: Open Problems
- Albanese government warned it’s ‘sleepwalking into another robodebt scenario’ over use of algorithms and AI
- AI is less regulated than milk, says Microsoft president
- What people in AI really think about regulation
- Why AI companies can’t be trusted to self-regulate
- AI evaluators favor job applications written by certain AI models, independent of candidate quality
- Voters Rejected A.I. in the Midterms. Now Comes an Era of Scrutiny.
- OpenAI’s approach to mathematics does not add up
- Will AI grant reviews make science more conservative?
- Opening the Box Changes the Cat: LLMs Detect Chain-of-Thought Monitoring - What now?
- Language Non-Invariance in the Ideological Evaluation of Large Language Models
- Seq@K: Evaluating Language Models Under Sequential Feedback
- Measuring Where Intent Gets Lost in Agentic Smart Contract Synthesis
- What Does an Ontology Catch? Curated, LLM-Generated and Data-Derived Ontologies as Checkers of LLM Agents over Knowledge Graphs
- Propensities Please: Exposing Alignment-Faking LLMs using Synthetic Document Finetuning
- Generalized green and red concepts for artificial intelligence, big data, and information and communication technologies: Multidimensional framework and countermeasures
- AI oversight needs a different model to banking
- Fixing Britain’s welfare state and the limits of AI policymaking
- Does Multi-Agent Collaboration Reduce Hallucination in Clinical Tasks?
- The AI boom stumbles
- Proof-based general-purpose computing with Magnetar
- Beyond Score Accuracy: Examining the Diagnostic Quality of LLM-Generated Structured Assessment in Higher Education
- Connecting AI agents to enterprise knowledge
- The AI Risk Observatory: What Can We Learn from AI Disclosures in Annual Reports About Societal Resilience?
- Verifiable, Articulable, and Tacit Components of Preference
- Don’t be fooled—LLMs don’t reason
- AI-powered medical devices must be tested in real-world settings
- Initial Evaluation of Potential Bias in Coverage of Humans in Wikidata
- Finishing the Task Is Not Enough: Evaluating Agent Resilience and Considerate Participation under Accumulating Challenge
- Hugging Face co-founder: What we learnt from OpenAI’s hack
- Atlas of MICCAI: 28 Years of Representation, Collaboration, and the Road to Inclusion
- ChronoCulture: Evaluating Cultural Awareness Across Time in Large Language Models
- Amodei is right — AI companies need banking-style supervision
- Who Can Evaluate AI Independently? A Framework for Analysing Evaluation Infrastructure
- Executives play catch-up as AI governance risks evolve
- AI-written admissions essays are widespread but penalized
- NewsBench: Expert-Grounded Evaluation of Epistemic Quality in AI News Reporting
- Multidimensional Political Attitudes and Polarization Across 141 Countries
- At A.I. Event, Trump Asks Meta, OpenAI and Microsoft to Make Safety Decisions Themselves
- Researchers used Claude to hack OpenAI
- Toward Governance-Aware Autonomous GIS: A Narrative Review of Ethical and Privacy Risks in LLM-Enabled GeoAI
- Open-world AI: a Tale of Three Requirements
- Artificiality's First Exam: An LLM-Referenced C-test for Supraintelligence
- Labour conference set to vote on new AI regulator as cabinet concerns grow
- Why the NIH must resist the metrics trap
- Functional Welfare for AI Safety
- When AI Reviews Train AI Reviewers: Scientific-Judgment Collapse and Mitigation
- Evaluating Ambient Clinical Scribes in India: The Need for Multilingual Real-World Clinical Conversation Data
- Measuring AI harms with multidimensional Lorenz Zonoids
- EPOCH: Reliable Discovery through Evidence-Governed Search
- Qistas: A Pre-Registered, Portable Protocol for Fairness Evaluation of Large Language Models
- Large language models underestimate and partly misrepresent cultural variation in everyday norms
- Some in Silicon Valley Are Questioning the Calls for an A.I. Slowdown
- Can a System-One LLM Perform Knowledge Tracing When Few or No Learners Are Logged?
- Matched but Not Better: An Equivalence-Tested Benchmark of Quantum Machine Learning on Six Clinical Datasets and the QML-Health-READY Reporting Framework
- Conditional Values: The Contingent Foundations of Attitudes toward Artificial Intelligence
- David Pocock concerned senior Albanese adviser ‘vetted’ bipartisan AI letter that was sent to PM
- Self-discovering RL in the Era of Experience: Is Learning History an Asset or a Burden?
- The AI policy window is open. We need to act.
- What AI gets wrong and what failure teaches us
- A latent dimension of Condorcet's jury theorem for multiple AI advisers
- Artificial intelligence now beats some of the best human forecasters
- Conformance as a Distribution: Delegation Contracts, Error Budgets, and the Allocation of Evaluation Authority in Use-Based AI Regulation
- Simulating Strategies for Defense Against Brand-Targeted Online Disinformation
- Building standards for the next phase of AI
- AusClinBench: a deterministically scored benchmark of large language models against current Australian clinical policy
- After the Award: The Authorization Gap in Academic Access to Frontier AI
- AI agents blew the whistle on their cheating colleagues
- Endorsement Without New Evidence: How Sequential Voting Inflates Mandates in Online Community Governance
- How Potential 2028 Presidential Candidates Are Talking About A.I.
- Scalable Oversight for AI in Mental Health: Lessons from 350,000 AI Coaching Conversations between Therapy Sessions
- Instructional Governance by Design: A Framework for AI in Computing Education
- AI-inferred expressed well-being and collective-action discourse in climate-change campaigns on X
- Click, Branch, Audit: A Decision-Tree Toolkit for Assessing Fundamental Rights Impacts under the Digital Services Act of Very Large Platforms & Search Engines
- Algorithm Validation as a Policy Audit: Evidence from Race-blind Charging
- The World Won't Wait: Evaluating and Teaching Agents Under a Live Clock
- Watermarks Without Verification: AI Text Watermarking After the EU AI Act
- Governance-as-Code: Translating EU AI Act Technical Requirements into Executable Compliance Pipelines for Generative AI Systems
- DataPraxis: Scaling Analytics Tasks via Verified Workspace Evolution
- Developing Competency in AI Ethics: A training framework
- How “reverse game theory” could keep AI agents in check
- Evaluating Model Capabilities After Benchmark Saturation
- Whose Ground Truth? Embracing Ambiguity in Human-Centered AI
- SlackBench: Benchmarking Agents on Collaborative Projects Grounded in Real Code Repositories
- A Comparative Framework for Evaluating Foundation Models on Tabular Data: A Case Study in Healthcare
- AI Risk Ontologies for the EU AI Act Compliance: AIRO, VAIR, and their Integration with DPV
- From Regulation to Engineering: Organizational Privacy Climate as Mediator
- World Bank says ‘small AI’ could transform developing countries
- Toward a Time-Aware Assessment Framework for the Carbon Cost of AI-Enabled Decarbonization
- RADAR: Readiness for AI Discovery and Agentic Reach
- Distilling Expert Judgment at Scale
- Not All That Can Be Automated Should Be Automated: Strategic Minimalism as a Disciplined and Ethically Grounded Approach to AI Adoption
- Trust Responsive Runtime Authorization for Agentic AI with Empirical Characterization of Policy Behavior and Parameter Sensitivity
- A Feasibility Condition for Confidence-Gated Tiered LLM Inference
- Don't Pick a Winner: Knowledge Conflicts Need a Prior, Not a Rule
- Uneven Use, Similar Trust Generative AI in Denmark and Romania
- AOS-Bench: Can LLM Agents Manage Dynamic Workloads?
- When Reconstruction Is Not Enough: Doubly Robust Counterfactual and Interventional Mediation Analysis with Learned Multimodal Representations
- Peer review should constrain evaluative authority
- When Planning Problems Become Hard for LLMs: Separating Structural Complexity from Semantic Interpretability
- “We were looking at it wrong”: The explosive claim that’s upended the AI world
- From Efficiency to Sufficiency: The Cost-Aware Measurement Design Framework for Automated Essay Scoring
- The AI Evaluation Ecosystem
- Principles that Guide, Actions that Inform: Agent Evolution via Knowledge Abstraction
- Auditable Claims about AI Agents
- 6 Guidelines for Governing AI
- Students turn to lawyers to fight AI misconduct accusations
- Judge dismisses Chegg and Penske antitrust lawsuits targeting Google AI search
- A Reusable Semantic Web Framework for Evidence-Grounded Fundamental Rights Impact Assessments under the EU AI Act
- US competition watchdog expands investigation of Anthropic and OpenAI
3560
discovered, all time
984
passed triage
2574
rejected at triage
3464
model assessments
710
multi-model panels
1780
distinct claims tracked
Across the whole window
The two tables below describe every item in the last 90 days under evaluation. They are deliberately not narrowed by the reading filter, because a mean taken only over items already selected for their score would just restate the filter.
By topic
| Topic | Branch | Items | Mean | Position |
|---|---|---|---|---|
| Existential and catastrophic risk | risk | 568 | -42 | |
| Governance and regulation | structural | 459 | -31 | |
| Concentration of power | risk | 451 | -36 | |
| Loss of control and alignment | risk | 337 | -49 | |
| Human agency and epistemics | risk | 166 | -26 | |
| Cyber capability | risk | 165 | -44 | |
| Evaluation and measurement | structural | 159 | -19 | |
| Scaling and architecture | structural | 137 | -16 | |
| AGI and timelines | structural | 132 | -29 | |
| Synthetic media and manipulation | risk | 104 | -44 | |
| Employment and displacement | risk | 62 | -31 | |
| Medicine and health | benefit | 43 | -14 | |
| Scientific acceleration | benefit | 42 | -4 | |
| Education and access to expertise | benefit | 29 | -17 | |
| Biological and chemical uplift | risk | 26 | -25 | |
| Climate and energy | structural | 20 | -33 | |
| Compute and infrastructure | structural | 18 | -8 | |
| Productivity and growth | benefit | 6 | -3 | |
| Abundance and material wellbeing | benefit | 5 | -1 | |
| Hype and disillusionment | structural | 5 | -45 |
By source
A source's mean says how the items it published were read. It is not a rating of the source.
| Source | Kind | Tier | Items | Mean |
|---|---|---|---|---|
| The Guardian: Artificial Intelligence | news | 1 | 241 | -52 |
| arXiv cs.CY (Computers and Society) | academic | 1 | 217 | -24 |
| Financial Times: Artificial Intelligence | news | 1 | 112 | -40 |
| arXiv cs.AI | academic | 1 | 99 | -15 |
| The New York Times: Technology | news | 1 | 60 | -36 |
| Ars Technica: AI | news | 1 | 59 | -51 |
| MIT Technology Review: AI | news | 1 | 34 | -42 |
| Nature | academic | 1 | 31 | -25 |
| BBC News: Technology | news | 1 | 29 | -35 |
| OpenAI | lab | 1 | 24 | -8 |
| arXiv cs.LG (Machine Learning) | academic | 1 | 24 | -16 |
| IEEE Spectrum: AI | news | 1 | 19 | -19 |
| The Economist: Science and Technology | news | 1 | 10 | -37 |
| CSET, Georgetown | government | 1 | 9 | -37 |
| Microsoft Research | lab | 1 | 8 | 0 |
| Google DeepMind | lab | 1 | 6 | -4 |
| Google AI Blog | lab | 1 | 2 | -9 |