Auto-Update News

We’re exploring the frontiers of AGI, prioritizing technical safety, proactive risk assessment, and collaboration with the AI community.

OpenAI is a non-profit artificial intelligence research company. Our goal is to advance digital intelligence in the way that is most likely to benefit humanity as a whole, unconstrained by a need to g

We (along with researchers from Berkeley and Stanford) are co-authors on today’s paper led by Google Brain researchers, Concrete Problems in AI Safety. The paper explores many research problems around

We’ve written a policy research paper identifying four strategies that can be used today to improve the likelihood of long-term industry cooperation on safety norms in AI: communicating risks and bene

We’re releasing Safety Gym, a suite of environments and tools for measuring progress towards reinforcement learning agents that respect safety constraints while training.

We’re introducing OpenAI Microscope, a collection of visualizations of every significant layer and neuron of eight vision “model organisms” which are often studied in interpretability. Microscope make

Now is a good time to start thinking about the governance of superintelligence—future AI systems dramatically more capable than even AGI.

We’re launching $10M in grants to support technical research towards the alignment and safety of superhuman AI systems, including weak-to-strong generalization, interpretability, scalable oversight, a

Advanced AI can transform biology and medicine—but also raises biosecurity risks. We’re proactively assessing capabilities and implementing safeguards to prevent misuse.

We (along with researchers from Berkeley and Stanford) are co-authors on today’s paper led by Google Brain researchers, Concrete Problems in AI Safety. The paper explores many research problems around

We are improving our AI systems’ ability to learn from human feedback and to assist humans at evaluating AI. Our goal is to build a sufficiently aligned AI system that can help us solve all other alig

Our mission is to ensure that artificial general intelligence—AI systems that are generally smarter than humans—benefits all of humanity.

Now is a good time to start thinking about the governance of superintelligence—future AI systems dramatically more capable than even AGI.

To support the safety of highly-capable AI systems, we are developing our approach to catastrophic risk preparedness, including building a Preparedness team and launching a challenge.

OpenAI is a non-profit artificial intelligence research company. Our goal is to advance digital intelligence in the way that is most likely to benefit humanity as a whole, unconstrained by a need to g

We describe our latest thinking in the hope of helping other AI developers address safety and misuse of deployed models.

Our mission is to ensure that artificial general intelligence—AI systems that are generally smarter than humans—benefits all of humanity.

Now is a good time to start thinking about the governance of superintelligence—future AI systems dramatically more capable than even AGI.

An Update for the UK AI Safety Summit

OpenAI and Anthropic share findings from a first-of-its-kind joint safety evaluation, testing each other’s models for misalignment, instruction following, hallucinations, jailbreaking, and more—highli

A Substack publication offering **news and analysis about the governance of transformative AI risks**, aimed at policymakers, tech enthusiasts, and engaged citizens.

We’re exploring the frontiers of AGI, prioritizing technical safety, proactive risk assessment, and collaboration with the AI community.

OpenAI is a non-profit artificial intelligence research company. Our goal is to advance digital intelligence in the way that is most likely to benefit humanity as a whole, unconstrained by a need to g

We’ve trained language models that are much better at following user intentions than GPT-3 while also making them more truthful and less toxic, using techniques developed through our alignment researc

We describe our latest thinking in the hope of helping other AI developers address safety and misuse of deployed models.

To support the safety of highly-capable AI systems, we are developing our approach to catastrophic risk preparedness, including building a Preparedness team and launching a challenge.

An Update for the UK AI Safety Summit

In this paper, we study the worst-case frontier risks of releasing gpt-oss. We introduce malicious fine-tuning (MFT), where we attempt to elicit maximum capabilities by fine-tuning gpt-oss to be as ca

OpenAI commits $7.5M to The Alignment Project to fund independent AI alignment research, strengthening global efforts to address AGI safety and security risks.

Our mission is to ensure that artificial general intelligence—AI systems that are generally smarter than humans—benefits all of humanity.

Now is a good time to start thinking about the governance of superintelligence—future AI systems dramatically more capable than even AGI.

Apollo Research and OpenAI developed evaluations for hidden misalignment (“scheming”) and found behaviors consistent with scheming in controlled tests across frontier models. The team shared concrete

OpenAI commits $7.5M to The Alignment Project to fund independent AI alignment research, strengthening global efforts to address AGI safety and security risks.

OpenAI’s mission is to build safe AI, and ensure AI’s benefits are as widely and evenly distributed as possible.

We’ve written a paper arguing that long-term AI safety research needs social scientists to ensure AI alignment algorithms succeed when actual humans are involved. Properly aligning advanced AI systems

Also, Anthropic Removes a Core Safety Commitment

An Update for the UK AI Safety Summit

To support the safety of highly-capable AI systems, we are developing our approach to catastrophic risk preparedness, including building a Preparedness team and launching a challenge.

Anthropic sues Trump administration in AI dispute with Pentagon, ‘Not built right the first time’ — Musk’s xAI is starting over again, again, Cascade of A.I. Fakes About War Wi

Also, Anthropic Removes a Core Safety Commitment

*Nature* (Published December 29, 2025) An editorial from *Nature* calling for global coordination on AI safety governance. The authorities in China are taking AI regulation extremely seriously, as ar

Our next iteration of the FSF sets out stronger security protocols on the path to AGI

OpenAI’s mission is to build safe AI, and ensure AI’s benefits are as widely and evenly distributed as possible.

We’ve contributed to a multi-stakeholder report by 58 co-authors at 30 organizations, including the Centre for the Future of Intelligence, Mila, Schwartz Reisman Institute for Technology and Society,

We are improving our AI systems’ ability to learn from human feedback and to assist humans at evaluating AI. Our goal is to build a sufficiently aligned AI system that can help us solve all other alig

Ensuring that AI systems are built, deployed, and used safely is critical to our mission.

To support the safety of highly-capable AI systems, we are developing our approach to catastrophic risk preparedness, including building a Preparedness team and launching a challenge.

We present a new research direction for superalignment, together with promising initial results: can we leverage the generalization properties of deep learning to control strong models with weak super

OpenAI, and our strategic partners, are thrilled about our shared vision for the Infrastructure of AGI. We are energized by the challenges we face and are excited by the prospect of partnering with fi