In this article
Four tensions are defining credentialing in 2026: Exam security, AI governance, workforce velocity, and candidate experience. Each one is real, each one is urgent, and none of them come with tidy, packaged answers.
Tension 1: Exam Security vs. a Professionalized Cheating Economy
Exam fraud isn’t what it used to be. What was once opportunistic has matured into a sophisticated, tech-enabled industry. Proxy testing services, AI-assisted cheating tools, deepfake identity fraud, and commercial brain-dump sites have all become permanent fixtures. Every credentialing program now has to plan around them.
The most resilient programs are responding by treating security as a multi-layered, “always on” system. In practice, that means:
- Convening cross-functional security groups—internal staff, delivery partners, proctoring vendors, and psychometric consultants—to share patterns, define risk appetite, and coordinate responses
- Leaning on long-term, multi-signal data forensics to disrupt organized fraud over time, not just flag individual bad actors
The scale of the problem is not abstract. In 2025 alone, Meazure Learning’s security team intervened in nearly 50,000 exam sessions due to security concerns and shut down 4,700 remote-control proxy sessions. For one client, our team identified 14 professional cheating rings within the first 2,000 exams administered. Fraud at this scale doesn’t disappear when it goes undetected; it just goes unaddressed.
What Credentialing Leaders Are Asking
The most effective approach is to design clear, transparent policies, communicate them early and often, and distribute security checks across the candidate journey so no single step feels punitive or surprising.
Proxy testing, AI-assisted cheating, deepfake identity fraud, and AI-enabled content harvesting coupled with commercial cheating services represent the sharpest end of the current threat landscape.
Tension 2: AI Efficiency vs. Human Accountability
AI is already embedded in the credentialing journey, including in job-task analysis, psychometric modeling, proctoring workflows, exam forensics, and item development. The question in 2026 isn’t whether to use it. It’s how to use it thoughtfully—in ways that accreditation bodies, regulators, and boards will find defensible.
The answer that most programs are landing on looks less like full automation and more like a structured middle path: AI assists, humans decide.
AI may conduct job task analysis, suggest item distractors, or flag anomalous sessions—but qualified professionals are still the ones who approve items, confirm scores, and sign off on exam decisions.
It’s worth noting, however, that credible gains in validity or cost savings from AI adoption are still emerging. Most programs are in early innings. The governance infrastructure being built now will determine how well programs can demonstrate defensibility later.
What Credentialing Leaders Are Asking
AI is evolving into a first-draft partner—helping SMEs generate items, scenarios, and variations faster—while humans retain authority over what enters the bank and how it aligns with the blueprint and validity evidence.
Credentialing bodies are adding explicit AI sections to their test security and data policies, defining acceptable uses, requiring vendors to support audits and documentation, and explaining to candidates how AI is—and is not—involved in exam development, scoring, and security.
Tension 3: Workforce Velocity vs. Credential Rigor
Are traditional, high-hurdle exam models still the best way to protect the public and move people into practice quickly?
That’s the question many credentialing programs are asking as labor shortages in health care, IT, and other high-stakes fields persist. The answer? “Not always”—followed by a serious effort to build faster, more flexible pathways to entry.
Three approaches are gaining traction:
- Performance-based testing—once considered logistically prohibitive—is becoming more scalable through automated scoring, offering a better way for candidates to demonstrate career readiness.
- Shadow scoring—where technology conducts scoring and experts audit a representative sample, typically around 5–10%—helps scale assessments while centering human judgment.
- Linear On-the-Fly Testing (LOFT) lets programs assemble equivalent forms on demand for more frequent sittings or modular content. Beyond flexibility, it also reduces exposure to leaked answer keys: When no two candidates see the same fixed form, stolen content loses much of its value.
What Credentialing Leaders Are Asking
Many programs experimenting with modular pathways are pairing module-level assessments with an integrative capstone exam—a holistic synthesis that candidates must pass to earn the full professional title—as a way to maintain rigor.
High-quality, job-relevant credentials can accelerate hiring—especially for non-degree and early-career workers. But many credentials deliver limited economic value, which puts pressure on programs to ensure their offerings genuinely align with what employers and the workforce actually need.
GenAI is shortening the shelf life of many technical skills while creating new AI-related skill demands. That’s pushing workers and employers toward faster, skills-first programs and credentials that can be updated more frequently.
Tension 4: Candidate-Centric Design vs. High-Stakes Standards
Candidate-centric design can sound like a euphemism for making things easier—softening the edges of a high-stakes process to reduce complaints and no-shows. But that’s not what the leading programs are doing, and it’s not what research supports.
When a candidate fails because the check-in process was confusing, the instructions were ambiguous, or the interface created unnecessary stress, the exam hasn’t measured competence; it’s measured a candidate’s ability to navigate a poorly designed system. That’s a validity and experience problem.
To remove that kind of friction in 2026, programs are mapping the full candidate journey—making expectations clearer, processes more intuitive, and avoidable stress less likely to contaminate results. The principle is straightforward, but the implication is significant: a well-designed candidate experience is a condition of measurement quality.
What Credentialing Leaders Are Asking
It means building policies, processes, and interfaces around real candidate needs and constraints—based on their feedback and behavior—while still meeting psychometric, legal, and security requirements. The goal is to make sure the exam measures what it’s supposed to measure, not a candidate’s ability to navigate the process.
Where Credentialing Leaders Go From Here
These four tensions are not problems to be “solved” and set aside. They are the new baseline for the industry.
Building a resilient program today means holding all four simultaneously—securing exams against a professionalized threat landscape, governing AI with documented human accountability, creating workforce pathways without sacrificing rigor, and designing candidate experiences that measure true competence.
If these are tensions your program is working through, we write about them regularly. Subscribe to our monthly LinkedIn newsletter for analysis and perspective as the credentialing landscape continues to evolve.