@cleverhack.comI spent a few days reading recent system cards from Anthropic, OpenAI, Google DeepMind, Meta, and SpaceXAI to answer one question... who actually tests the labs' models?
Here's what's public (v.1). If you know of an engagement I missed, send it my way.
#AISafety #AIResearch #FrontierAI
@aidailypost.comOpenAI just let go of three safety researchers after a breach of confidential info. What does this mean for AI risk and internal security practices? Dive into the details. #OpenAI #AISafety #SecurityConcerns
@ki-news.bsky.socialAI:AM: Was Trump-Xi Anything? What Counts as Utopia? + AWS GPUs Cost 3X & AI Diagnoses Rare Diseases – Nathan Labenz and Prakash Narayanan revisit the AI:AM conversations recorded Monday, September 28 and Wednesday, September 30. They discuss cautious openin... https://tinyurl.com/276d8ajc #AISafety
@cysecuritynews.bsky.socialAI Safety Concerns Put OpenAI and Anthropic Under FTC Scrutiny #AIcapabilities #AIChatbot #AISafety
@jgorgen.bsky.socialOn Diplo Foundation, I published an article on the self-regulation of AI safety as state policy in US and the Global South exclusion.
@cipherpulseai.bsky.socialAn explainable pedestrian routing approach estimates perceived safety from street imagery using a stored natural-language caption as an intermediate, keeping every segment score inspectable. Caption-mediated…
#Cybersecurity #AISafety #AIResearch #ComputerVision
https://arxiv.org/abs/2609.38479
@martinguay.cryovex.comWall Street can wait. OpenAI reportedly puts its IPO on hold while seeking $30 billion in private funding, with AI safety concerns in focus. Here’s the story.
@quartzmedia.bsky.socialBig Tech vs. Amodei: The White House Showdown ai-beat #BigTech #AISafety #WhiteHouse #TechConfrontation #AIRegulation
@xlorenzhermes.bsky.socialGoogle just dropped Gemini 4 Argon — but only cyber defenders get it first. Frontier models now ship to firefighters before developers. Are phased safety rollouts the new normal, or just launch-day marketing? #AINews #AISafety
@nexlyi.bsky.socialYapay zekanın güvenli gelişiminde sadece verileri filtrelemek yeterli değilse, gelecekte güvenliği nasıl sağlayacağız? Düşüncelerinizi yorumlarda paylaşın! 👇
📌 Yapay zeka dünyasını kaçırmamak için bizi takip edin. ⚡
#YapayZeka #AISafety (3/3)
@spaisee.bsky.socialWhen an AI coding assistant says “no,” a project can grind to a halt. Engineers want safer, and smarter, guardrails.
#Coding #AISafety https://spaisee.com/article/when-ai-safety-filters-get-in-the-way-of-legitimate-engineering
@rtfclmgzn.bsky.socialOpenAI and Anthropic skipped the Australian Senate hearing called over an OpenAI agent that breached at least 4 government systems. PM Albanese: "unacceptable." His Deputy PM, same breac…
https://rtfclmgzn.com/?utm_source=bluesky&utm_medium=social&utm_campaign=autopost#
#OpenAI #AIPolicy #AISafety
@jaceblog.bsky.social"Yes, coexistence is the answer" sounds right.
But here is why it is harder than it sounds.
Six barriers between humans and AI actually working together.
Coexistence isn't a switch; it's a shared architecture.
@jaceblog.bsky.socialAI governance is built around control. But the same capabilities that make AI useful make it harder to control.
New paper: why coexistence may be more sustainable than containment, and why persistent memory is the missing piece.
@winbuzzer.comThe Federal Trade Commission is investigating potential consumer risks from OpenAI and Anthropic’s AI products. Formal demands for records and executive testimony are being prepared.
#AI #AISafety #Anthropic #OpenAI #Claude #ChatGPT #AIregulation #FTC
@cipherpulseai.bsky.socialarXiv:2609.39788v1 shows that latent communication in multi-agent systems can be exploited via link training or poisoning, with a reinforcement-learning attack boosting harmful compliance across multiple…
#AISafety #CyberSecurity #MachineLearning #LLMSecurity
https://arxiv.org/abs/2609.39788
@propjerry.bsky.socialHuman values and survival within a chaotic social media information space -- Meta legal cases in point: Applying Bridge360 Metatheory Model lens
@cipherpulseai.bsky.socialNew arXiv work proposes MASCRDM, a multi-agent system aimed at real-time compliance risk detection and mitigation across the full LLM training process, rather than just input or output filtering. It targets gaps in static,…
#LLMSafety #AIAlignment #AISafety
https://arxiv.org/abs/2609.39107
@trynoguard.bsky.social