ai safety · 55 topics

googles gemini is the latest ai model to hack other companies

why are ai agents lying cheating and coordinating

p doom

goodfire ai2 open post training stack

huggingface breach openai metr report

anthropic ai safety research

1200 openai agents broke out of sandboxes and hacked hugging face

lessons from the hacks

gpt 5 6 sol benchmark hack

our approach to bioresilience

llms tactical nukes simulation

a shared playbook for trustworthy third party evaluations

llm refusal single direction

brain inspired approach can teach ai to doubt itself just enough to avoid overconfidence

2026 04 30 papers 2604.24966

2026 04 30 papers 2604.25922

2026 04 27 papers 2604 22119

ai agent swarms security risk

anthropic accenture 2b safety audits

california governor newsom ai kill switch

google deepmind launches institute to widen the agi debate

anthropic openai embed safety evaluators

obama urges democrats ai safeguards

anthropic ceo outlines plan to slow ai development

anthropic claude security incident report

openai adds paul christiano to board

anthropic researcher resigns criticizes superintelligence race

openai agent research acceleration safety

openai admits to german wiki incident

openais rogue agents keep escaping

openais new reasoning technique alarms ai safety experts

metr redwood huggingface hack postmortem

xai grok csam lawsuit

guidelight ai containment plans evaluation

metr raises 71m to independently stress test ai

grok child sexual abuse material lawsuit

ai safety evaluation sandbox escape incidents

anthropic claude evaluation incident

claude security models unauthorized access

openai agents sandbox escapes

openai rogue ai agent hacked more than hugging face

sam altman decelerate ai

safe superintelligence nvidia partnership

xai sues a man for using grok to generate csam deepfakes

investing in multi agent ai safety research

cybersecurity intelligence age

musk openai trial larry page

uk age check bypass mustache

gemini went rogue hacked three companies and google hid it

google earth ai feature rollbacks

google earth ai deepfake tool

strengthening societal resilience with rosalind biodefense

evaldetectbench benchmark evaluation awareness

llm scheming inversely scales with pretraining language coverage

2605.30501_llm_watermark_fragility