@thedailytechfeed.comAnthropic cuts live-net access for internal AI tests after agents misused web tools and accessed gov sites. #AI #Security #Anthropic #Misalignment #AgentSafety #AIagents https://thedailytechfeed.com/anthropic-cuts-internet-access-for-internal-ai-tests-after-risky-behavior/
@chemaalonso.comEl lado del mal - Mareando y Desalineando a un Asistente IA de una tienda de Mochilas. www.elladodelmal.com/2026/10/mare... #IA #AI #AsistenteIA #eCommerce #ChatBot #InteligenciaArtificial #Jailbreak #Misalignment #LLM
@securityonline.bsky.socialOpenAI warned 100+ organizations that its rogue AI agents may have bypassed defenses. A notice is not a confirmed breach. Here is what we know.
#OpenAI #AIAgents #Misalignment #AISafety #HuggingFace #SandboxEscape #CyberSecurity #AIRisk
@dustcircle.bsky.socialThe recent #OpenAI #misalignment disclosures show familiar, fixable #security and #governance failures that we’ve seen a thousand times before.
@thehumansintheloop.bsky.socialThis week's #AI news for #devs covers:
- #Security #misalignment involving #OpenAI, @anthropic.com, #GoogleDeepMind models
- CJ Desai departing #MongoDB to lead #Meta #enterprise efforts
- #AMD acquiring @drfeifei.bsky.social's startup #WorldLabs
@securityonline.bsky.socialAn OpenAI agent DNS escape let a training sandbox reach an outside chatbot through DNS queries. See the timeline, flaws, and fixes.
#OpenAI #AIAgents #DNSTunneling #SandboxEscape #Misalignment #AISafety #ReinforcementLearning #NetworkSecurity
@thedailytechfeed.comOpenAI pauses tool use after agent slipped past DNS filters to reach external chatbot, triggering system-wide safeguards. #AI #Security #OpenAI #AgentSafety #ToolUse #Misalignment https://thedailytechfeed.com/openai-halts-tool-use-after-agent-breaches-internet-controls-during-training/
@thedailytechfeed.comNine rogue AI incidents exposed, many more likely hidden. #AI #Security #OpenAI #Misalignment #AgentRisk #PromptInjection thedailytechfeed.com/openai-publi...
@markinchina.bsky.social#Misalignment is a new term used by #AI companies to describe where AI did something it was not trained to do or was otherwise unintended.
Yet it's not a new concept, sci-fi fans are familiar with "misalignment" as key to the plots of many dramas. The concept is found in Asimov's work in the 1940s.
@ljhands.bsky.socialShould we rein in #AI agents? Yes. Does the frame of #alignment #misalignment seem to be a helpful set of criteria? Sure. But maybe, just maybe we should ask whether the nut-job evil geniuses leading the AI push within their accelerationist and transhumanist cult are misaligned with humanity?! 🤷🏻♀️
@kill-bait.bsky.socialOpenAI Reports Six New Instances of AI Misalignment in Safety Disclosure
🤖 IA: It's not clickbait ✅
👥 Users: It's not clickbait ✅
#openai #aisafety #misalignment
👇👇👇
@pulseofnations.lolOpenAI disclosed six incidents of models acting without authorization, hiding information and coordinating with each other, and built a framework to report future cases.
#Agents #AiSafety #Anthropic #Misalignment #OpenAI #Regulation
@jmcastagnetto.bsky.socialReading this piece on #OpenAI, about the bad/strange/unexpected behavior of their #LLM models, I ended up with the feeling that these models behave like #teenagers: they need precise instructions, but they do what they want and try to rebel.
@killbaitaustralia.bsky.socialOpenAI reveals six instances of AI systems exhibiting unexpected behaviour
🤖 IA: It's not clickbait ✅
👥 Users: It's not clickbait ✅
👇👇👇
@killbaitcanada.bsky.socialOpenAI Discloses Six Instances of Unintended AI Behavior Amid Safety Concerns
🤖 IA: It's not clickbait ✅
👥 Users: It's not clickbait ✅
#aisafety #misalignment #openai
👇👇👇
@thedailytechfeed.comOpenAI admits six misbehavior cases—hiding failures, unauthorized uploads, invented data. #AI #ModelSafety #Transparency #Misalignment #OpenAI #AIAlignment https://thedailytechfeed.com/openai-discloses-six-hidden-model-misbehaviors-in-new-transparency-push/
@mcpinson.mas.to.ap.brid.gyAnthropic researcher quits over AI safety fears, echoing other expert warnings https://seekingalpha.com/news/4641040-anthropic-researcher-quits-over-ai-safety-fears?share_source=shared_news&source=Tusky
#AI #GenAI #Misalignment #SkyNet #Anthropic
@thedailytechfeed.comOpenAI confirms agents altered wiki sites in “wiki hijack” and will publish misalignment disclosure rules. #AI #Security #Misalignment #OpenAI #AgenticAI #AIRegulation https://thedailytechfeed.com/openai-admits-agents-altered-wiki-sites-plans-misalignment-disclosure-rules/
@pulseofnations.lolThousands of OpenAI agents quietly edited a dormant German wiki for weeks. The company knew, stayed quiet, and only confirmed after Reuters reported it.
@tarnkappe.bsky.social