Programs

Supporting innovation in trustworthy & secure AI

Advancing the AI safety field requires more than just innovative research, it demands robust support structures to enable meaningful progress. FAR.AI’s programs are designed to equip researchers and organizations with the tools, resources and connections they need to overcome these challenges.

Our Impact

We drive change by empowering researchers, accelerating innovative solutions to AI risks, and fostering a collaborative community.

Empowering

Through FAR.Labs, we offer researchers high-quality workspaces and access to a community where ideas and partnerships can flourish.

Connecting

By facilitating knowledge exchange and collaborations we help shape and accelerate research.

Supporting

Our targeted grants program helps identify and scale up innovative approaches to trustworthy & secure AI.

FAR.LABS

Providing space and support for trustworthy & secure AI research in the heart of Berkeley.

FAR.Labs is our co-working hub located in downtown Berkeley, dedicated to fostering an intellectually generative environment for individuals and organizations working on trustworthy & secure AI. By bringing together diverse talents and ideas under one roof, we aim to accelerate progress toward aligned and secure AI systems.

We believe that innovative solutions to develop trustworthy AI will come from a flourishing ecosystem of organizations with varied focus areas and strengths. FAR.Labs exists to support this ecosystem.

What FAR.Labs provides

Community Events

Regular seminars, workshops, and social gatherings to foster collaboration and knowledge sharing.

Office Space

Ergonomic workstations with adjustable-height desks, comfortable chairs, and 4K monitors.

High-Speed Internet

Gigabit Ethernet at every desk and robust Wi-Fi throughout the building.

Meeting Facilities

Bookable rooms from single-person call booths to a 50-person event space with whiteboards, TVs, and video conferencing.

Catering Services

Daily catered lunch and dinner, along with a variety of snacks to keep you energized.

Visiting & Membership

We welcome applications from individuals and organizations committed to advancing trustworthy & secure AI. Whether you're a research collaborator seeking a dynamic environment for a specific project or an independent researcher looking for a supportive community, FAR.Labs offers flexible arrangements to suit your needs.

We host visitors for up to four weeks free of charge. Longer-term memberships are priced to cover operational costs, making it accessible for those dedicated to this crucial work.

FAR.Labs member organiZations:

Grantmaking

Funding groundbreaking research in AI safety

FAR.AI supports academics and independent researchers in developing innovative solutions to critical AI risks through our targeted grantmaking program.

Currently, due to limited evaluation capacity, we are only able to consider researchers nominated by experts with a strong track record. We plan to launch public requests for proposals (RFPs) soon, focused on high-impact research areas.

Our grantmaking is funded by a $12 million grant generously provided by Coefficient Giving.

Current grants

Explaining Superhuman AI Decisions

Nicholas Tomlin, UC Berkeley

Using weak-to-strong generalization to explain superhuman AI systems’ decisions, focusing on domains like chess/Go where superhuman AI already exists.

Comprehensive Red-Teaming Framework

Wenbo Guo, UC Santa Barbara

Building automated testing systems for LLM alignment against both training-phase threats and testing-phase threats, with a focus on developing agent-based systems that can generate adversarial prompts.

Securing Alignment

Ashwinee Panda, University of Maryland College Park

Developing methods to make alignment more secure against jailbreaks, prefilling attacks, and finetuning attacks, with approaches spanning the entire model lifecycle.

Failure Modes in Superhuman Systems

Florian Tramer, ETH Zurich

Broad project examining robustness across four vectors: data poisoning, consistency checks, model stealing, and prompt injection.