Implementation Therapy: The Five Stages of Grief in AI Adoption

In one of the most talked-about sessions at the Talkpush Americas Summit 2025, Raquel and Luis —part of the global implementation team— explained why adopting artificial intelligence in recruitment is not just a tech initiative, but a deeply human process.

Instead of focusing on features, metrics or interface updates, they offered a different lens: AI adoption follows the same five stages of grief described by Elisabeth Kübler-Ross. And after implementing AI in companies across multiple countries and cultures, they’ve seen the same emotional curve repeat itself over and over.

This editorial summary captures the essence of their talk, designed for Talent Acquisition teams navigating change today.

Denial: “This won’t work here”

The first reaction when a company adopts AI is denial.

Typical comments include:

• “Our process is different.”
  
• “I don’t think the bot will understand what we do.”
  
• “This sounds too good to be true.”
  

Even when leadership is committed to digital transformation, operational teams often compare themselves to other companies and assume they’re the exception. According to the speakers, during the early years of the AI wave, 42% of AI projects stalled at this exact stage—not due to technical failures, but because the organization never crossed the emotional barrier required for change.

The takeaway: if a solution works for companies with similar challenges, there’s a strong chance it can work for yours too.

Anger: “Why are they changing this?”

After denial comes fear—and fear often turns into frustration.

Fear of losing control.
Fear of losing your job.
Fear of not understanding the new system.

During kickoff meetings, recruiters ask how far the bot will go, what it will automate and how it will impact their daily work. When the first mistake appears or a message isn’t perfect, some use it as proof that the technology “doesn’t work.”

Raquel and Luis emphasized that this is not intentional resistance. It’s simply the human brain reacting to the unknown. The way forward is education: asking questions, exploring the tool, admitting what we don’t know yet and giving the system time to improve with use.

Negotiation: “Yes… but slowly”

Once the anger settles, many teams enter the stage of “yes, but…”

• “Yes, it works, but let’s extend the pilot.”
  
• “We’ve seen results, but let’s try it only with one area.”
  
• “Let’s wait a bit so we don’t risk the operation.”
  

This becomes an emotional tug-of-war between moving forward and holding back. To prevent this stage from dragging on forever, the speakers recommend setting a clear commitment period —30, 60 or 90 days— during which everyone uses the tool as if it were already the long-term solution. Only after that window should results be reviewed.

Consistency, not perfection, is what gets teams through negotiation.

Depression: “I don’t even care anymore”

This is the most dangerous and, according to the speakers, the most common stage.

Depression in AI adoption doesn’t look like sadness.
It looks like indifference.

People stop logging in.
Stop reporting issues.
Stop trying new features.
Stop asking questions.
Stop caring.

Recruiters return to manual calls because “it’s what they know,” teams stop championing the initiative, and transformation quietly starts to crumble. In this stage, 82% of companies abandon their AI project—not because the technology fails, but because the organization loses emotional momentum.

Talking about this stage openly and recognizing it early is the only way to move past it.

Acceptance and Growth: “Now it makes sense”

Once a company pushes through indifference, it reaches a turning point.

Teams begin proposing improvements.
The AI model adapts more accurately to real-world workflows.
Recruiters trust automated processes.
AI expands into referrals, internal mobility, screening, interviewing and more.

A year ago, only 8% of organizations successfully scaled AI across the entire recruiting operation. That number is rising, and many companies attending the Summit are already part of that progress.

This is the moment when AI stops feeling like an add-on and becomes operational infrastructure.

A new mindset: training AI like an intern

The talk closed with a simple but powerful analogy:

Adopting AI is like receiving a new intern.

If you don’t train them, correct them or explain your processes, they won’t help you.
But if you coach them daily—fixing messages, adjusting flows, improving prompts—they learn quickly, and every correction compounds over time.

Teams that succeed with AI are the ones who embrace this coaching mindset: curious, patient and aware that every small improvement scales across thousands of candidate interactions.