Just one year ago, our Salesforce Architect survey revealed a profession in transition. AI adoption was already high, with 88.9% of architects using AI tools in some capacity, but confidence was lagging. Nearly a third (32.6%) felt behind their peers, concerns around cost and security were widespread, and Agentforce confidence sat at just 54.4%.
Fast forward to 2026, and the picture looks noticeably different.
According to the 2026 SF Ben Architect Survey, AI adoption has surged to 96.7%, confidence levels have improved across multiple measures, and architects are beginning to move from simply using AI to integrating it into the way they work. While barriers still exist, the data suggests that Salesforce Architects are becoming increasingly comfortable with AI as part of their day-to-day toolkit.
The question is no longer whether architects are adopting AI. It’s how quickly they are advancing with it.
AI Adoption Is Now Nearly Universal
Last year’s survey showed that 88.9% of Salesforce Architects were already using AI regularly or on an ad hoc basis. At the time, that felt impressive. In 2026, that figure has climbed to 96.7%.
Even more notably, almost two-thirds of respondents (63.6%) now report using AI daily or regularly, while only 3.2% say they do not use AI at all.
For a profession that has historically approached new technologies with careful consideration, this level of adoption signals a significant shift. AI is not viewed as an emerging trend anymore. It’s becoming a standard component of the architect’s toolkit.
The AI Skills Gap Is Narrowing
Perhaps the most encouraging finding from this year’s data is the improvement in how architects perceive their own AI capabilities.
In 2025, 32.6% of respondents felt behind their peers when it came to AI knowledge and skills. Only 17.8% believed they were ahead.
This year, the percentage of architects who feel behind has fallen to 25.5%.
Meanwhile:
- 50.8% feel they are on par with their peers.
- 20.6% feel ahead of their peers.
- Just 3.1% have not yet started building their AI skill set.
Combined, 71.4% of architects now feel they are either on par with or ahead of their peers.
That may not sound revolutionary at first glance, but it represents a meaningful shift in confidence over a relatively short period of time.
Last year, much of the discussion centred around a growing knowledge gap. Today, the data suggests architects are actively closing it.
Productivity Remains the Biggest Win
One of the strongest consistencies between both surveys is the value architects are seeing from AI. In 2025, 66.1% reported productivity gains from AI. In 2026, increased productivity remains the most commonly cited benefit, recognized by 63.9% of respondents.
While the headline number has dipped slightly, the broader picture reveals a maturing set of use cases.
This year, architects highlighted:
- Increased productivity (63.9%)
- Creating documentation (54.1%)
- Finding information faster (53.5%)
- Brainstorming (52.3%)
- Greater efficiency (50.7%)
- Faster learning (48.1%)
Compared with last year’s findings, architects appear to be using AI for a much wider range of activities.
In 2025, the conversation was largely centred around productivity and information retrieval. In 2026, architects are increasingly leveraging AI for documentation, learning, ideation, and solution development.
This suggests a growing level of sophistication in how AI is being applied across the architecture lifecycle.
Agentforce Confidence Is Climbing
One of the most significant improvements can be seen in confidence around Agentforce.
Last year, architects reported an average confidence score of 54.4% when architecting Agentforce-powered solutions. This year, that figure has risen to 61.7%.
Advanced architects report even higher confidence levels at 63.5%, while intermediate architects average 62.4%. Junior architects trail slightly at 57.5%, but still exceed last year’s overall average.
While confidence levels are not yet where they might be for mature Salesforce technologies, the upward trend is difficult to ignore.
As more organizations move from pilots into production deployments, architects are gaining valuable real-world experience that simply wasn’t available a year ago.
The AI Tool Landscape Has Changed
The AI tool ecosystem has also evolved significantly.
Last year, ChatGPT dominated the landscape, with an impressive 86.3% adoption rate among architects. Google Gemini followed at 43.5%.
In 2026, ChatGPT remains the most widely used tool at 62.7%, but the market has become considerably more diversified.
Architects now report using:
- ChatGPT (62.7%)
- Claude (52.1%)
- Google Gemini (45.9%)
- Agentforce (33.4%)
- GitHub Copilot (20.6%)
- Agentforce Vibes (20.3%)
- Einstein AI features (17.2%)
- Cursor (13.2%)
- Microsoft Azure AI services (11.6%)
The decline in ChatGPT’s share is not necessarily a sign of reduced usage. Rather, it suggests architects are broadening their AI stacks and selecting specialized tools for different tasks.
The emergence of Claude, Cursor, Copilot, and Agentforce as mainstream tools reflects a market that is rapidly maturing.
Trust Has Replaced Cost as the Biggest Challenge
One of the most interesting shifts appears in the barriers architects face.
In 2025, cost was the clear number one concern, cited by 21.3% of respondents. Accuracy followed at 18.3%, with skills and knowledge at 17%.
This year, the barriers are much more evenly distributed:
- Trust (19.3%)
- Skills and knowledge (19.0%)
- Accuracy (17.6%)
- Company buy-in (17.3%)
- Cost (16.1%)
- Capacity and resources (10.8%)
The drop in cost concerns may indicate that organizations have become more willing to invest in AI initiatives, or that architects are finding value that justifies the spend.
Instead, trust has emerged as the leading challenge.
This evolution makes sense. As architects move beyond experimentation and begin incorporating AI into business-critical processes, the quality, reliability, and governance of AI outputs become increasingly important.
The conversation has shifted from “Can we afford AI?” to “Can we trust it?”
Final Thoughts
Looking back at last year’s findings, the dominant theme was potential. Architects could see the benefits of AI. They were experimenting with tools. They recognized the productivity gains. But many still felt behind, uncertain, or underprepared.
The 2026 data tells a different story.
Adoption has increased. Confidence has improved. Agentforce expertise is growing. The skills gap is narrowing. Architects are using a wider range of tools and applying AI to a broader set of tasks.
Challenges remain, particularly around trust, governance, and organizational buy-in. But the profession appears to be moving beyond the early adoption phase and into something more meaningful.
Salesforce Architects are no longer asking whether AI has a place in their work. They’re figuring out how to use it better than everyone else.
