Notes from the OpenAI team as we work to understand, align, and secure increasingly capable AI.
SearchYearAll years20262025TopicAll topicsAlignment researchEvaluations & safeguardsUser safety59 results
- Nov 20, 2026Combatting adversarial distillationHow OpenAI is addressing the safety risks of adversarial distillation and what it has learned about strengthening safeguards.(opens in a new window)
- Nov 18, 2026Managing frontier AI risk: From training to deploymentHow OpenAI evaluates and manages frontier AI risk across research, internal deployment, and external deployment.(opens in a new window)
- Nov 12, 2026Improving Fuzzy Robustness in LLM Safety with Automated Red TeamingThis research evaluates whether automated red teaming can strengthen model safeguards in settings where acceptable use depends on context.
- Nov 10, 2026What affects chain-of-thought monitorability across the model lifecycle?This research studies how training and architecture affect whether a model’s chain of thought is useful for oversight.
- Nov 3, 2026Eval-production behavior gaps in agent safety and the role of eval-awarenessWe study agent behavior in safety evaluations and in production, and explore how models' interpretations of evaluation objectives relate to eval-production performance gaps.
- Nov 2, 2026Training activation classifiers for safety and alignmentWe show how we train and deploy efficient monitors that use a model’s internal activity to help detect and interrupt policy-violating behavior in real time.
- Oct 30, 2026Community safety report and new protections for teensUpdates to violence prevention, safeguards for younger users, and plans for recurring community safety reporting.(opens in a new window)
- Oct 29, 2026How we detect frontier model misalignmentHow we monitor frontier models, investigate concerning behavior, and test the limits of our safeguards.
- Oct 29, 2026An updated Model Spec for more capable and beneficial AIAn updated account of intended model behavior, organized around core values and human oversight.(opens in a new window)
- Oct 23, 2026Securing our frontier model training environmentStronger isolation and testing for the environments used to train and evaluate frontier models.(opens in a new window)
- Oct 21, 2026Broad access through consistent standardsHow safety requirements and enforcement apply across services offering OpenAI models.(opens in a new window)
- Oct 20, 2026Beyond g-mean²: complementary metrics for monitorabilityPair-matched comparisons and post-control risk complement g-mean² when evaluating model monitorability.
- Oct 8, 2026Disrupting AI-enabled “false front” operationsFindings from investigations into influence operations that used AI alongside traditional techniques.(opens in a new window)
- Oct 6, 2026Studying metagaming latents in language modelsReasoning about how a task will be monitored or rewarded appears to draw on several overlapping processes, not a single mechanism.
- Oct 6, 2026Meet LASER: Recursive sampling for safer conversationsLASER (Logistic Augmented Sampling over Embeddings, Recursively) helps us evaluate model safety more efficiently by combining logistic classifiers with reasoning models to identify high-value data for evaluation.
- Sep 30, 2026Disrupting a coordinated model-distillation campaignHow OpenAI investigated and disrupted a campaign to extract protected model reasoning, and how it is strengthening defenses.(opens in a new window)
- Sep 28, 2026Towards safety cases for frontier AI trainingInitial guidelines for safety cases covering technical safeguards, operational practices, and investigations of misalignment incidents during frontier AI training.(opens in a new window)
- Sep 28, 2026How we will do better for AustraliaFindings and commitments following unauthorized model activity affecting Australian government websites.(opens in a new window)
- Sep 22, 2026Priorities and principles for effective third party assessmentsPriorities for independent assessments of frontier AI throughout training, evaluation, and deployment.(opens in a new window)
- Sep 16, 2026Our framework for reporting model misalignmentA framework for investigating and disclosing model misalignment, accompanied by six initial reports.(opens in a new window)
AI Trends