Setting the Pace
Where the social sector is placing its bets in the Year of the Horse.
2026 is the Year of the Horse, symbolizing momentum, power, and forward motion. That seems about right for AI. The technology isn’t slowing down, but this feels like a moment where direction matters just as much as speed.
The tools are stronger. The stakes are clearer. And the social impact sector has moved past asking whether AI belongs. The real questions now are how it shows up, who it serves, and what standards it’s held to.
I asked some friends and leaders close to the work where they think AI for humanity is headed in 2026. The same word kept showing up in their answers: choice. Choices about trust and guardrails, openness and scale, and when AI should move fast versus when it really shouldn’t.
Saddle up. 2026 might be the year we find out whether we’re steering AI in the right direction — or just holding on for the ride.
+ At Fast Forward, we’re adding to our own horsepower in 2026. We’re expanding our team to support more AI-powered nonprofits doing this work — if you’re interested in joining us, check out our nine open roles.
Jared Chung, Founder, CareerVillage.org
“In 2026, the social impact and AI ecosystem will move from abstract vision to concrete proof. This will be a year defined by unified and deepened collaboration across the social sector, clearer and more cohesive definitions of success (ie: impact measurement), and a narrowing of the field as products are tested against real outcomes rather than ambition alone. We’ll start to see which tools actually earn trust because they work — for practitioners, for institutions, and most importantly for the people they’re meant to serve. At the same time, I expect growing pressure on frontier AI companies to meet the bar for quality, safety, and usefulness that social sector leaders are deliberately setting together — aligning through coalitions to establish shared guardrails and protect users in a rapidly shifting landscape.”
Afua Bruce, Co-Author, The Tech That Comes Next and Principal, ANB Advisory Group
“In 2026, institutions shape responsible AI by investing in three critical areas: teams, infrastructure, and narrative. First, institutions must fund teams with expertise in both technical development and socio-economic impact and support them in creating datasets, AI tools, and evaluation frameworks. Second, they must sustain the infrastructure — human capital and technical capacity — that allows other organizations to build on proven tools. Third, institutions must champion responsible AI through clear communication, demonstrating that well-designed AI can deliver a positive impact.
Institutional choices influence whose knowledge powers AI systems, which impacts matter, and what solutions receive sustained funding—ultimately deciding whether AI reinforces inequity or advances the public good.”
Raffi Krikorian, Chief Technology Officer, Mozilla
“The tipping point for open-source AI won’t come from ideology — it’ll come from economics and convenience. We’re already seeing rapid movement toward small models (1B-8B parameters) running on edge devices, API prices rising, and enterprises feeling trapped in closed ecosystems. At the same time, geopolitical pressures (e.g. export controls, regional restrictions, sovereignty demands) are pushing organizations toward self-hosted, auditable systems. 2026 is the year open-source AI stops being the principled choice and starts being the rational one. The question isn’t whether openness wins… it’s whether we build the infrastructure to make it easy before the defaults harden.”
+ Raffi expands on this thinking — and what Mozilla is building — here.
Vanessa Parli, Managing Director of Programs and External Engagement, Stanford HAI
“In 2026, I expect scientific discovery to accelerate meaningfully as researchers are increasingly augmented by AI, unlocking new breakthroughs in healthcare, sustainability and beyond. More of these advances will reach the public as companies embed the newest models in user friendly interfaces and tools like Claude for Healthcare. This shift creates real opportunity for nonprofits and mission-driven organizations to adopt powerful capabilities that were once out of reach, helping close persistent gaps in resources and expertise. Benchmarks, too, will evolve, moving away from abstract performance scores toward domain-specific evaluations that tie models to real-world outcomes in areas like healthcare, education and law. At the same time, I expect a recognition that while AI can be transformative for certain tasks and workflows, it is ill-suited for others. This clarity will drive more thoughtful deployment, focusing AI where it genuinely adds value. Finally, sovereign AI will gain momentum, as governments seek greater independence from large frontier labs by building their own models and infrastructure, with an emphasis on culturally aware design and development.”
Rey Faustino, Founder and CEO, One Degree
“In 2025, we saw a lot of experimentation using AI for eligibility and enrollment in social services and public benefits, but that also revealed massive limitations, especially around reliability, explainability, and trust given the non-deterministic nature of generative AI. When you’re enrolling someone into something as critical as public benefits, ‘close enough’ isn’t acceptable. The system has to be precise. In 2026, I expect we’ll see those challenges meaningfully addressed through better guardrails, hybrid systems that blend deterministic eligibility engines with AI to continuously improve them, and purpose-built models for government. And with that shift, I predict we’ll see the first U.S. state adopt a generative AI–powered enrollment solution (something we’re building at One Degree!), so we’ll see the move from AI being a pilot to real public infrastructure.”
Gemma Turon and Miquel Duran-Frigola, Co-Founders, Ersilia
“In 2026 we will see a new wave of agentic AI systems for medical research that go beyond task automation and accelerating routine work, becoming active contributors to scientific discovery. In the next few months, we will see breakthroughs in drug discovery not only accelerated by AI, but made possible thanks to it. Agentic AI will become a true companion for medical researchers, amplifying our horizon.”
Scott Kleper, AI Advisor, Fast Forward
“2026 will be the year that AI is front-and-center as a collaborator throughout the day. Rather than serve as a tool that is employed in specific use cases, teams will have adaptable AIs helping them map out their days, keeping projects on the rails, and serving as sounding boards when creativity strikes. The risk is using AI in a generic way that feels hollow. Nonprofit organizations that train their AI collaborators to deeply understand their mission will punch above their weight because the AI is natively speaking their language, not offering generic help.”
Caroline Spears, Executive Director, Climate Cabinet
“There’s an old saying that goes something like this: ‘You may not be interested in politics, but politics is interested in you.’ This is hitting the tech world first hand as backlash grows across the country to AI and data centers — fueled by anger over questions around rising electricity rates, water access, and double standards that carve the industry out from clean energy commitments. 2026 will be the year that AI companies will run headlong into the physical and legal constraints of the existing electricity infrastructure. In the process, they risk sparking political and voter dissent over rising rates. But there’s another path: AI could be a partner in clean energy contracts, lowering electricity prices, and a commitment to community. Advancements in LLMs and NLP could be used to solve big problems. 2026 will be a year of choice.”
Quick Bytes
Other Sector Stories
2026 will be built on the learnings of 2025 — which makes the Collective Intelligence Project’s 2025 Global Dialogues Index essential reading. Remember back in August when I shared their findings about AI consciousness and trust? They’ve now compiled a year’s worth of data from 6K+ people across 70 countries into one comprehensive report.
The Patrick J. McGovern Foundation built Grant Guardian, a Claude-powered tool that standardizes financial due diligence for grantmakers. It’s the kind of practical infrastructure 2026 needs, already adopted by 100+ philanthropies and built to free up time for the work that matters.
Fast Forward alum CareerVillage.org is helping steer the AI for Career Development Coalition, which brings together 80+ organizations across education, workforce development, and philanthropy to set shared standards for AI in career guidance. Building trust, transparency, and impact measurement as a coalition instead of going it alone? That’s how you move a sector forward.
APN Opportunities and Funding News
Mozilla is accepting nominations for its 2026 Fellows program, supporting up to 10 technologists, researchers, and advocates building better technology futures. Nominations close January 30.
The StepForward: Essential Innovation Challenge is seeking practical solutions helping young people build the skills and pathways needed for an AI-powered future. Selected projects will share a $1.2M prize pool to expand AI literacy and connect learning to real career opportunities. Apply by February 6.
Let’s Talk
I am living and breathing AI for humanity these days. If you are too, let’s talk!