How We Vet AI-Capable Candidates at CrossBridge (Our Full Process)
AI-capable candidates are not hard to find because nobody uses AI.
They are hard to find because many people can name tools, but fewer can show good judgment.
At CrossBridge, we look for candidates who can use AI to produce better work, communicate clearly, and operate inside a real business workflow. Tool familiarity matters, but it is not enough.
SHRM has reported that 51% of organizations use AI in recruitment, which means both employers and candidates are changing how hiring works. SHRM, How AI Is Reshaping Talent Acquisition in 2026
Our process is built to separate surface-level AI usage from practical capability.
Stage 1: Role clarity
Before screening candidates, we define what the role needs.
That includes:
Business goal
Required skills
Nice-to-have skills
Tools used by the team
Communication expectations
Time-zone needs
Dealbreakers
This matters because “AI-capable” means different things in different roles.
For an engineer, it may mean using AI to support coding, testing, and debugging. For a support hire, it may mean summarizing tickets and improving knowledge-base content. For an operator, it may mean automating reporting or documenting workflows.
We do not look for generic AI talent. We look for role-specific capability.
Stage 2: Communication screen
Remote and nearshore work depends on communication.
We evaluate whether the candidate can:
Explain work clearly
Ask useful questions
Summarize decisions
Share blockers early
Write in a way customers or teammates can understand
Receive feedback
Stay organized across tools
This is not about sounding polished. It is about creating clarity.
If communication is weak during the process, it is unlikely to become strong after hiring.
Stage 3: AI workflow review
Next, we ask candidates how they use AI.
We want specific examples, not tool lists.
Useful prompts include:
“Walk me through a workflow where AI helped you produce better work.”
“What did you use the tool for?”
“How did you check the output?”
“What did you change before sharing the final version?”
“When would you avoid using AI?”
“How do you protect confidential information?”
Strong candidates can explain the before and after. They can describe the problem, the tool, the review step, and the result.
Weak candidates usually stop at: “I use ChatGPT.”
Stage 4: Practical assessment
For many roles, we use a short role-specific exercise.
The goal is not free work. The goal is to see how the candidate thinks.
Examples:
An engineer reviews a small code sample and explains risks.
A support candidate improves a customer response.
An operations candidate maps a broken process.
A recruiter screens a candidate profile against a scorecard.
A marketing candidate turns a messy brief into a clear outline.
We pay attention to how they use AI, but also how they edit, validate, and explain the work.
What we look for
Across roles, strong AI-capable candidates show:
Clear thinking
Practical tool use
Quality control
Data awareness
Good writing
Ownership
Curiosity
Security awareness
Humility about AI limitations
Ability to improve a workflow
The best candidates do not treat AI as magic. They treat it as a tool that still requires judgment.
What raises a red flag
We are careful with candidates who:
Cannot explain how they use AI
Submit work that feels generic
Do not validate AI outputs
Ignore privacy concerns
Use jargon instead of clear explanations
Avoid accountability
Overstate their expertise
Struggle to explain tradeoffs
AI can make weak answers sound smoother. That is why the process needs specific follow-up questions.
SHRM Labs has also noted that assessment design must adapt as candidates can use AI tools during live assessments. A 2026 report found that 59% of hiring managers suspect candidates of using AI tools during live assessments. SHRM Labs, Assessment design and AI
The answer is not paranoia. It is better assessment design.
Why this process works
This process works because it looks at real work behavior.
We are not asking whether someone has heard of ChatGPT, Claude, Cursor, Notion AI, or Make. We are asking whether they can use tools responsibly inside a business setting.
That is what companies actually need.
The practical takeaway
AI-capable hiring should be specific, practical, and grounded in judgment.
Define the role. Screen for communication. Ask for real workflows. Use short assessments. Look for validation habits.
The best candidates are not the loudest about AI. They are the ones who use it carefully to produce clearer, faster, better work.