OpenAI Faces Multi-State Investigation Over AI Safety and User Harm Concerns
A coalition of US state attorneys general is investigating OpenAI amid growing concerns over AI safety, accountability, and user harm.
#Safety/Alignment
609 articles tagged with Safety/Alignment
US government issues AI recall, forcing Anthropic to pull Fable 6 and Mythos 5 over jailbreak panic
The fast-moving world of artificial intelligence just hit its first massive regulatory brick wall, and it is unlike anything we have seen before. In an unprecedented move, Anthropi...
UCLA Health launches center to evaluate AI safety and implementation in health care
UCLA Health launched the INOVAi Center on June 11, 2026, to test AI safety in clinical care. Trials show AI scribes cut physician note-writing time and reduce exhaustion.
This also implies the Alignment Problem of keeping #AI to human goals is intractable without constant human input, as mo...
This also implies the Alignment Problem of keeping #AI to human goals is intractable without constant human input, as model collapse will always pull towards the unaligned minima o...
Microsoft Signs MoU to Collaborate With Singapore Regulator on AI Safety, Security
Microsoft has signed a memorandum of understanding to collaborate with Singapore's Infocomm Media Development Authority on artificial intelligence safety and security, the latter s...
bingook/bingo: Bingo - AI-powered Red Team Terminal (DeepSeek/Claude/GPT/GLM)
Bingo - AI-powered Red Team Terminal (DeepSeek/Claude/GPT/GLM)
Former xAI engineer sues company over AI safety concerns, claims he was laid off for speaking up
A former xAI engineer has accused Elon Musk's artificial intelligence company of dismissing him after he repeatedly warned about the risks posed by Grok. The lawsuit paints a pictu...
đź“° A former xAI engineer has filed a lawsuit against the company and SpaceX, alleging he was fired for raising AI safety ...
đź“° A former xAI engineer has filed a lawsuit against the company and SpaceX, alleging he was fired for raising AI safety concerns about Grok days before SpaceX's historic IPO.đź”— http...
🤖 Guided Model Alignment Frameworks Gain Traction in AI ResearchResearchers are increasingly focusing on inference time ...
🤖 Guided Model Alignment Frameworks Gain Traction in AI ResearchResearchers are increasingly focusing on inference time alignment methods to improve the performance of large langua...
Why Your Next.js SaaS Needs a Production AI Agent Guardrail Architecture, Not Just a Prompt
I've spent the last year building production AI pipelines for SaaS platforms. The prompts were solid....
Former XAI Engineer Sues Elon Musks SpaceX For Firing Over AI Safety Concerns
A former engineer at Elon Musk’s xAI who now heads a think tank focused on AI safety filed a lawsuit claiming he was fired from the SpaceX subsidiary for raising concerns about the...
🤖 AI red teaming comes of age📝 When Ram Shankar Siva Kumar launched Microsoft’s AI red team in 2019, the discipline bare...
🤖 AI red teaming comes of age📝 When Ram Shankar Siva Kumar launched Microsoft’s AI red team in 2019, the discipline barely existed. “The running jok...https://www.csoonline.com/art...
Grammar-Constrained Decoding Can Jailbreak LLMs into Generating Malicious Code
Large Language Models (LLMs) are increasingly used for code generation, raising concerns that they may be misused to produce malicious code. Meanwhile, Grammar-Constrained Decoding...
AI Safety Expert: These People Will Only Survive Till 2030
Make yourself and your family AI-scam proof, step by step → https://neuralnutshell.com Roman Yampolsky, who coined the term ...
Center for AI Safety Names Former Robinhood and Meta Executive Rochelle Nadhiri as Head of Public Engagement
SAN FRANCISCO--(BUSINESS WIRE)--The Center for AI Safety (CAIS), a nonprofit focused on reducing societal-scale risks from artificial intelligence, today announced the appointment ...
The Role of Feedback Alignment in Self-Distillation
Conditioning a language model on additional context, such as feedback on a previous attempt, typically improves its response. Self-distillation trains the model to retain this impr...
IDEAL: In-DEpth ALignment Makes A Discrete Representation AutoEncoder
Built on pretrained vision foundation models (VFMs), representation autoencoders (RAEs) have recently emerged as a promising approach for constructing semantically rich latent spac...
Survey reveals 80% would jailbreak their Kindle before letting Amazon winAndroid Authity readers want control over their...
Survey reveals 80% would jailbreak their Kindle before letting Amazon winAndroid Authity readers want control over their pre-2012 Kindle devices.https://www.androidauthority.com/su...
Emergent Misalignment Can Be Induced by Sycophancy and Reversed via Alignment Gating
Prior work has shown that fine-tuning large language models on malicious or incorrect outputs in narrow domains can induce broad misalignment and harmful behavior, a phenomenon kno...
BenSyc: Benchmarking Conversational Sycophancy and Human Alignment in LLMs for Bengali Contexts
Large language models (LLMs) increasingly participate in emotionally sensitive social conversations, where responses may shift from balanced support toward excessive validation or ...
AI in SDLC: Why I Stopped Optimizing for Code Generation and Started Optimizing for Alignment
Over the past few months I built an AI-assisted delivery framework — not to write code faster, but to...
MIT symposium examines AI alignment, education and the limits of machine reasoning
MIT researchers gathered April 30 to examine what happens when AI logic doesn't match human reasoning. A central warning: replacing institutions with AI before understanding how th...
Evals Are Alignment Enforcement: Why Your Safety Strategy Needs Runtime Checks
The Argument The AI safety conversation is dominated by two camps: the alignment...
MaskAlign: Token-Subset Representation Alignment for Efficient Diffusion Training
Representation alignment with pretrained vision models has recently shown strong potential for accelerating diffusion transformer training. By aligning intermediate diffusion featu...