Forum Hits of 2025

At OpenAI, we’re building safe AGI that benefits all of humanity. We look for people who are inspired by this mission and ready to tackle big challenges

Jack Stubbs, from the Intelligence and Investigations team, described how his group disrupts organized scam networks while also empowering the public to use ChatGPT as a personal safety tool. He emphasized that most scammers are not inventing new methods but using AI to scale old tricks more efficiently, and that OpenAI has disrupted operations in Cambodia, Myanmar, and Nigeria. Stubbs underscored both the human and financial toll of scams, citing $12 billion in reported U.S. losses last year and even teen suicides linked to sextortion. His team uses a “ping–zing–sting” framework to map scam patterns and has found AI involved at every stage. Importantly, he highlighted that millions of people already use ChatGPT to identify scams, with three times more scam-detection interactions than scammer interactions, and noted growing demand for free, accessible AI safety tools. Stubbs closed by stressing transparency through public reports, partnerships with groups like AARP, and collaboration across tech and civil society to ensure AI strengthens safety, security, and shared benefits for all.

Łukasz Kaiser’s OpenAI Forum talk, “Learning Powerful Models: From Transformers to Reasoners and Beyond” offered a research-focused but deeply values-aligned reflection on how AI is evolving from data-hungry systems toward reasoning models that learn more efficiently and safely. The framing he used emphasized safety, learnability, and human-like reasoning. He consistently underscored that making AI more learnable from less data and more computationally powerful ensures that progress in AI remains beneficial, efficient, and accessible to all, rather than concentrated among a few actors.

This discussion between OpenAI's Chief Economist Ronnie Chatterji and LinkedIn’s Chief Economist Karin Kimbrough explores the current and future impact of generative AI on the labor market, highlighting macroeconomic dynamics, evolving skill demands, and global disparities in AI adoption. Kimbrough offers labor market insights based on LinkedIn data, while Chatterji connects these to policy, infrastructure, and human-centered innovation themes promoted by OpenAI.

Dr. Anton Maximov delivered a compelling presentation on how artificial intelligence is revolutionizing neuroscience by enabling the analysis of complex brain structures at an unprecedented scale and speed. His work uses AI-driven tools—particularly convolutional neural networks integrated with 3D electron microscopy—to uncover the nanoscale architecture of long-term memory. This research, previously impossible with manual techniques, showcases how AI transforms the pace, precision, and possibility of scientific discovery. Dr. Maximov also reflected on how AI is streamlining everyday scientific tasks, democratizing hypothesis generation, and setting the stage for more dynamic, scalable, and biologically inspired AI systems. His talk illustrated a future where AI not only accelerates research but redefines how science is conceived, conducted, and shared.

This forum conversation showcased OpenAI’s deep alignment between its internal culture and its global mission: to build AGI that benefits everyone. Joaquin’s career story—spanning safety research, a personal return to hands-on coding as an intern, and his unexpected move into recruiting—served as a powerful narrative about humility, mission-alignment, and human-centered innovation. The conversation emphasized OpenAI’s commitment to democratizing AI access, building infrastructure for economic transformation, and advancing AI that reflects democratic values, not autocratic control.

The conversation, featuring educator and California teacher of the year, Casey Cuny showcased how ChatGPT is transforming education. Cuny emphasized AI literacy as a moral imperative, framing AI as an opportunity. He spoke about democratizing AI benefits, boosting productivity, expanding educational access, and reinforcing the importance of ethical, human-centered use. The dialogue also highlighted teacher adoption, policy considerations, and workforce readiness—areas where democratic AI values, infrastructure investment, and shared benefits intersect.

The OpenAI Forum hosted educator and data artist Greg Niemeyer from UC Berkeley for a talk on how AI is transforming learning, teaching, and thinking. Building on OpenAI’s mission to ensure broadly distributed benefits from AI, Niemeyer introduced a “Minus AI, Plus AI, Times AI” framework: minus AI for intentionally tech-free, embodied learning; plus AI for transparent, dialectical collaboration with AI; and times AI for AI as a medium that restructures knowledge itself. He proposed a cognitive insight formula—C = Q × T × K where meaningful learning depends on the quality of questions, the strength of trust, and the richness of the knowledge base, emphasizing that if trust collapses, learning outcomes collapse as well. Throughout the talk, he shared concrete classroom experiments showing how AI can either de-skill students or spark creative divergence and multiplayer learning when used thoughtfully and transparently. He closed by urging educators and learners not to choose one mode, but to move wisely among minus, plus, and times AI to keep curiosity, meaning, and our shared “we” at the center of education in the age of AI.

During the virtual event on December 3rd, Prof. Terence Tao and OpenAI's Mark Chen and James Donovan engaged in a deep discussion on the intersection of AI and mathematics. They explored how AI models, particularly new reasoning models, could enhance traditional mathematical problem-solving and potentially transform mathematical research. The speakers discussed the integration of AI into various scientific fields, emphasizing AI's role in accelerating discovery and innovation. Key topics included the challenges of AI in understanding and contributing to complex mathematical proofs, the evolving nature of mathematical research with AI integration, and the future of collaboration between AI and human mathematicians. The conversation highlighted both the potential and the current limitations of AI in advancing mathematical sciences.

Renowned mathematician Terence Tao and OpenAI Chief Research Officer Mark Chen were joined by OpenAI’s VP of Science Kevin Weil and such luminaries as Caltech mathematician Sergei Gukov, UCSB physicist Nathaniel Craig, Stanford’s Eva Silverstein, Lance Dixon of SLAC National Accelerator Laboratory, UCLA’s Zvi Bern, Wahid Bhimji of Lawrence Berkeley National Laboratory and NERSC, University of Wisconsin physicist Kyle Cranmer, and OpenAI’s Alex Lupsasca and James Donovan for talks, panels, and public discussion.

The content in this presentation covers subjects that may be distressing, including technology-facilitated child sexual abuse and exploitation. No actual or depicted child sexual abuse material is contained. Attendees are encouraged to practice self-care/wellness as needed. At this OpenAI Forum event, Natalie Cone hosted a conversation with Julie Cordua, CEO of Thorn, and Chelsea Carlson, who leads child safety efforts across OpenAI’s products, focused on protecting children in digital spaces. They described how online harms have evolved over the last decade, including increased grooming, sextortion, and the rise of synthetic or AI-generated child sexual abuse material. Julie explained Thorn’s approach across research with youth, technical innovation, and building tools that help platforms and law enforcement detect abuse, triage cases, and find victims faster. Both speakers emphasized there is no single fix, and that meaningful progress requires safety-by-design, clean training data, strong guardrails, scalable detection, and clear pathways for reporting and response. They also underscored the mental toll on investigators and moderators and discussed how AI can reduce unnecessary exposure by grouping, prioritizing, and filtering sensitive material without replacing human judgment. During Q&A, they highlighted the importance of real-time, multimodal, and contextual detection, and shared practical guidance for parents centered on engagement, literacy, and keeping open lines of communication with kids. The session closed with a call for deeper collaboration among nonprofits, tech companies, and governments to improve capacity, transparency, and cross-border coordination to keep children safer online.