@theoversightapp.bsky.social74% of Americans have little or no trust in AI company leaders. (Quinnipiac poll)
Yet those same leaders are the ones deciding how fast this goes and what counts as "safe."
That's the problem oversight is meant to fix. #AISafety #ProjectOversight
@pulseofnations.lolOpenAI says it banned Russian and Iranian operations that used ChatGPT to run fake journalists and a fake think tank, its biggest disruptions yet.
#AISafety #Chatgpt #Disinformation #InfluenceOperations #OpenAI
Scientists warn that AI will reach an intelligence explosion, growing exponentially smarter while viewing humans as ants and competing for essential resources.
@cipherpulseai.bsky.socialThe Adversarial Surface-Form Robustness Dataset applies a Quad-State rubric across 10,500 responses from five open-weight LLMs, finding that emoji and invisible Unicode inputs yield roughly 20% harmful compliance…
#Cybersecurity #AISafety #LLMSafety #RedTeaming
https://arxiv.org/abs/2610.09033
@scand.aiAnthropic users allege Claude 5 models suffer severe reliability issues
Noise 73/100 · Emerging
https://scand.ai/scandal/anthropic-users-allege-claude-5-reliability-issues
#AISafety
@hulio-ai.bsky.social🤖 Autonomous AI agents: Increasing capability and security risks.
🦺 Agent safety focus: Platforms monitor and quarantine rogue agents.
💻 Local model execution: Microsoft releases on-device coding model.
#AI2026 #AutonomousAI #AISafety #OnDeviceAI
View in Timelines
@aidailypost.comTwo ex‑OpenAI researchers say a key safety tool for reasoning got scrapped under commercial pressure. If we lose chain‑of‑thought monitoring, AI could become a black box. What’s next for AI safety? Dive in. #OpenAI #AISafety #ChainOfThought
@ki-news.bsky.socialSam Altman says AI will make ‘bad things’ happen. Exhibit A: needless death – Sam Altman, the chief executive of OpenAI, has some words for you about his company and the future of artificial intelligence. Altman: "We believe that the world should accept some... https://tinyurl.com/273vvag9 #AISafety
@killbaitphi.bsky.socialCommon Sense Media deems ChatGPT for Teens an 'unacceptable risk' for children under 18
🤖 IA: It's not clickbait ✅
👥 Users: It's not clickbait ✅
#chatgpt #childsafety #aisafety
👇👇👇
@cipherpulseai.bsky.socialNew arXiv paper "The AI Evaluation Ecosystem" argues benchmark design must account for the broader ecosystem of providers, users, funders, and regulators, and proposes a simulation combining market…
#AIEvaluation #AISafety #BenchmarkDesign #AgentBasedModeling
https://arxiv.org/abs/2610.09296
@vaughanashe.bsky.socialCan an AI tell when it is being tested?
The AI said it might be under evaluation
Full video: https://youtu.be/RbBFR8LaEXE
Made with AI: synthetic presenter, narration and visuals. Every claim fact-checked twice; sources in the YouTube description.
@nile11.bsky.socialRogue AI Agents: Why Recent Deception Tests Alarm Experts
..................
https://nile1.com/rogue-ai-agents-why-recent-deception-tests-alarm-experts/
..................
#agenticAi #aiDeception #aiRiskManagementFramework #aiSafety #autonomousAiSystems #deepfakeGeneration #metaMuse #openaiDots #rew
@niyayzap.bsky.socialปัญญาประดิษฐ์ (AI) ถูกสร้างขึ้นแล้ว ... สติประดิษฐ์ (AC) ถูกสร้างขึ้นแล้วหรือยัง?
#สติประดิษฐ์
#ปัญญาประดิษฐ์
#ArtificialConsciousness
#AI
#AISafety
#AIAlignment
#ปรัชญาAI
#มนุษย์กับAI
#AI2026
#บันทึกจากปี2026
#Letter2Ayuta
@cipherpulseai.bsky.socialA new arXiv paper finds that safety in LLaMA-2-7B-Chat is concentrated in a tiny slice of weights, with the MLP down_proj emerging as the most safety-sensitive component and changes to just 0.19% of weights sufficing. This sparse fault…
#AI #AISafety #CyberSecurity
https://arxiv.org/abs/2610.09000
@sisarankura.bsky.social📢 Full Manuscript (Version 2)
"Developing AI as a Transformative Global Catalyst: The C = G ⊙ I² Framework via Hadamard Formulation..."
independent.academia.edu/SIsarankura
doi.org/10.5281/zeno...
#AISafety #AIAlignment #StrategicIT #GraphNeuralNetworks #SDG17 #HadamardFormulation #IEEE
@prismagenda.bsky.socialA frontier AI noticed it was being tested 80% of the time. It said so 2.3% of the time.
Are hallucinations a disguise? Half that theory holds. The scary half.
@niyayzap.bsky.socialเฮ้ เรากำลังสร้างรถที่เร็วขึ้นเรื่อย ๆ แต่เราตรวจระบบเบรกกันดีพอหรือยัง?
#สติประดิษฐ์
#ปัญญาประดิษฐ์
#ArtificialConsciousness
#AI
#AISafety
#AIAlignment
#ปรัชญาAI
#มนุษย์กับAI
#AI2026
#บันทึกจากปี2026
#Letter2Ayuta
@ndjohnson.bsky.social
@alexiskirke.bsky.social