Nila Kartika Wati
Nonprofits and non-governmental organizations across the globe are integrating generative artificial intelligence into their daily operations at a pace that far exceeds the development of internal governance, ethical guardrails, and comprehensive staff training. According to the "2026 State of Nonprofit AI: Adoption and Governance Report," a landmark study conducted through a partnership between NTEN and The Bridgespan Group, the social sector is currently navigating a period of unprecedented technological transition characterized by high enthusiasm but significant structural fragility.
The report, which synthesized data from nearly 1,000 respondents surveyed between late April and mid-June, paints a picture of a sector in the midst of a "quiet revolution." While public discourse often focuses on the theoretical doomsday risks of AI or the massive environmental toll of the data centers powering these models, the reality on the ground for nonprofits is one of rapid, often uncoordinated implementation. Of the 917 organizations surveyed, a staggering 98 percent reported using AI in some capacity, with 61 percent utilizing the technology in an official, organized manner. However, the infrastructure required to manage this technology—ranging from dedicated budgets to ethical use policies—remains dangerously underdeveloped.
The Infrastructure-Governance Paradox
The central finding of the NTEN and Bridgespan research is what analysts call the "implementation-governance gap." While the adoption of AI tools—including generative AI for content creation and AI-powered features within standard office software—has become nearly universal, the institutional frameworks to support them have lagged behind.
The report identified five critical pillars of AI readiness: dedicated budgets, staff training, physical and cloud infrastructure, data organization, and formal knowledge of responsible use. In every category, the survey results indicated a deficit. Most notably, 57 percent of nonprofit executives reported that their organizations have no dedicated budget for AI. Furthermore, 58 percent of organizations lack an AI roadmap, and only 8 percent have a fully realized strategic plan for the technology’s integration.
This lack of formal planning extends to the very data that fuels AI systems. Less than half of the surveyed executives (40 percent) have established written guidance regarding what types of organizational or constituent data can be safely used with AI tools. Similarly, only 38 percent have issued formal policies on the responsible or ethical use of the technology. This vacuum of policy means that a significant portion of AI usage in the sector is occurring by default rather than by design, raising concerns about data privacy, security, and mission alignment.
A Chronology of Rapid Transition
To understand the current state of the sector, it is necessary to look at the timeline of AI’s arrival in the nonprofit world. Between 2023 and 2025, the release of high-functioning large language models (LLMs) triggered a wave of "experimental adoption." During this period, many nonprofits began using AI for low-stakes tasks such as drafting social media posts, summarizing meeting notes, or generating initial grant outlines.
By early 2026, the technology moved from the periphery to the core of nonprofit operations. The NTEN/Bridgespan survey reflects this shift, showing that 45 percent of respondents now use AI for work at least daily. Another 29 percent use it regularly, at least once a week. This rapid escalation from "novelty" to "necessity" has left IT departments and executive boards struggling to catch up. The report suggests that the sector has bypassed the traditional "pilot phase" of technology adoption, moving straight into team-wide or fully integrated use before the risks were fully assessed.
The Executive-Staff Disconnect
One of the more revealing aspects of the 2026 report is the divergence in how AI adoption is perceived at different levels of the organizational hierarchy. The data suggests that executives and frontline staff are experiencing the AI transition through two very different lenses.
Executives are generally more comfortable with the technology, with 67 percent reporting a high level of comfort compared to 55 percent of staff. Executives are also more likely to use AI daily (50 percent) than their subordinates (38 percent). Perhaps most tellingly, staff members frequently describe AI adoption as something happening "to and around them." While 63 percent of organizations report that AI decisions are made at the executive level, only 1 percent of respondents indicated that AI implementation was a collective or all-staff decision.
This top-down approach has led to a sense of alienation among some staff members, who view AI as a pressure driven by external trends rather than an intentional tool for mission advancement. While 25 percent of staff noted the presence of an internal AI council or ethics committee, the majority felt that adoption was reactive rather than collaborative.
Shadow AI and the Risks of Informal Use
A byproduct of the governance gap is the rise of "shadow use"—the informal, unofficial use of AI tools outside of organizational guidance. The report found that 53 percent of all respondents engage in shadow use. In a surprising twist, executives were found to engage in shadow use at a higher rate (57 percent) than staff (49 percent).
This phenomenon presents a significant risk to the sector. When employees or leaders use personal accounts or unvetted AI tools to process organizational data, they bypass security protocols and privacy protections. In the nonprofit sector, where organizations often handle sensitive information regarding vulnerable populations, the stakes of such "off-the-books" technology use are particularly high. The report suggests that without clear, official pathways for AI use, employees will continue to find their own solutions, further complicating the task of data management and ethical oversight.
Ethical Concerns and Environmental Impact
The 2026 report also highlights a growing awareness of the externalities associated with AI. While data privacy remains the top concern for executives (55 percent), staff members are increasingly focused on the environmental costs of the technology. A significant 66 percent of staff respondents cited the environmental impact of AI—specifically the massive energy and water consumption of data centers—as a major concern. This was the highest response rate for any concern category measured among staff.
Field-level risks also weigh heavily on the minds of sector leaders. Approximately 47 percent of executives expressed concern that AI will widen the "digital divide," creating a larger gap between populations with access to AI-enabled services and those without. Other cited concerns include algorithmic bias, the potential for AI to erode community trust, and the displacement of human labor.
Despite these anxieties, only 6 percent of respondents reported a specific AI-related "harm" or incident in the previous 12 months. However, the authors of the report caution against complacency. They suggest that the low rate of reported harm may be an "iceberg effect," where incidents go undetected or undisclosed because organizations lack the monitoring systems to identify when an AI tool has hallucinated, leaked data, or produced biased outcomes.
Strategic Frameworks for a Sustainable Future
To address these challenges, the report points to the "Choosing Your AI Path" framework developed by The Bridgespan Group. This model encourages nonprofit leaders to move beyond reactive adoption and instead make strategic choices based on their organization’s specific mission, resources, and risk tolerance.
The path forward, according to the report, requires a shift in financial priorities. While 49 percent of executives anticipate that AI spending will grow significantly in the coming year, that spending must be balanced between software licenses and the "human infrastructure" of training and governance. Currently, only 32 percent of nonprofit staff have received formal AI training. Even in well-resourced organizations, nearly 60 percent of the workforce remains untrained in the responsible use of the technology.
Industry analysts suggest that the nonprofit sector is at a crossroads. If organizations can close the gap between adoption and governance, AI has the potential to serve as a powerful force for the public good, streamlining operations and expanding the reach of social services. However, if the current trend of unguided implementation continues, the sector risks exacerbating existing disparities and compromising the trust of the communities it serves.
The 2026 State of Nonprofit AI report serves as a definitive call to action. It suggests that the "fix" for the current instability is not to stop using AI, but to start governing it. By committing resources to clear guidance, ethical frameworks, and comprehensive staff training, nonprofits can ensure that their technological evolution remains aligned with their core values of equity and service. As the technology continues to mature, the ability of the social sector to manage its risks will be just as important as its ability to harness its potential.
