As AI continues to pervade the world of recruitment, one trend has been steadily rising: AI-driven video interviews. Promoted as faster, cheaper and more “objective,” these tools promise to screen candidates at scale, without human involvement.
However, as more organisations adopt them, a growing number are beginning to ask a critical question:
"Are AI video interviews actually helping us hire better people - or are they quietly undermining the quality of our decisions?"
At Leonid, our view is clear. While technology can improve efficiencies in recruitment operations, removing humans from candidate assessment is a mistake and particularly for senior, specialist and high-impact roles.
Leonid’s bespoke video interview service LeonidLive does not – and never will – use AI. It is a live video interview, conducted only by humans and reviewed only by humans.
The problem with AI video interviews
AI video interviews are often positioned as neutral and efficient. In practice, they introduce a number of risks that organisations are only beginning to understand.
1. They are optimised for performance, not potential
AI interview tools typically assess how candidates speak, structure answers or present on camera. What they cannot reliably assess is:
- Judgment
- Depth of experience
- Ethical reasoning
- Contextual decision-making
- Communication skills
- Personality and cultural fit
This creates a bias toward candidates who are comfortable performing for an algorithm, rather than those best equipped to succeed in complex, real-world roles.
Strong operators, thoughtful leaders and technically brilliant specialists are often filtered out: not because they lack capability, but because they don’t “interview well” for a machine.
2. They strip away context
Recruitment decisions are rarely binary. Experience, career choices and communication styles only make sense in context.
AI interviews struggle to account for:
- Non-linear career paths
- Sector-specific nuance
- Cultural or regional communication differences
- The why behind decisions
Human interviewers naturally explore these areas. Algorithms do not.
3. They risk candidate disengagement and brand damage
For many candidates - particularly senior professionals - AI interviews feel impersonal and transactional. Instead of feeling assessed, they feel processed. This can result in candidates feeing disengaged, damage to an employer’s brand and talent being lost to competitors who do offer a more ‘human’ experience.
In a market where top candidates have choice, experience matters.
4. They create a false sense of objectivity
AI interview tools are often marketed as unbiased. In reality, they are trained on historical data and behavioural patterns, which means they can replicate existing biases, not eliminate them. Crucially, they also lack the ability to explain why a candidate was rejected in meaningful, defensible terms, thereby creating risk in regulated or senior hiring contexts.
Why LeonidLive Is different
LeonidLive was built 7 years ago, before the rise of AI. It was originally created to bring CVs to life and make the hiring process more efficient, as hiring managers are able to quickly assess the communication skills and personalities of shortlisted candidates. All of this still very much holds true today, but now there is an added benefit: it’s an antidote to AI.
Rather than replacing human judgment, LeonidLive enhances it.
It is a live, human-led interview and assessment service, delivered by experienced consultants who understand both the role and the market.
LeonidLive interviews are conducted by experienced consultants who can:
- Probe experience in depth
- Challenge assumptions
- Explore judgment, ethics and decision-making
- Adapt questioning in real time
This produces insight that no AI tool can replicate.
Designed for high-impact roles
LeonidLive is particularly effective for:
- Risk, audit, compliance, legal, ESG and governance roles
- Senior and board-facing positions
- Roles requiring judgment, credibility and influence
- Cross-functional or hybrid mandates
These are roles where how someone thinks matters as much as what they’ve done.
Better decisions, better outcomes
Because LeonidLive assessments are contextual, nuanced and human, clients gain:
- Clearer evidence to support hiring decisions
- Reduced mis-hires
- Higher candidate engagement
- Greater confidence at board and stakeholder level
It’s not about speed for its own sake: it’s about making the right decision the first time.
Technology should support humans – not replace them
At Leonid, we’re certainly not anti-technology. We have adopted a wide suite of tools which help us deliver the best possible service to our clients: we believe in using data, insight and innovation to improve hiring outcomes.
But when it comes to assessing people - their judgment, potential and fit - humans still outperform machines.
Final Thought
In an increasingly automated recruitment landscape, the organisations that stand out will be those that retain human judgment where it matters most.
AI video interviews may save time. However, human-led assessment builds stronger teams.
LeonidLive is able to offer the best of both worlds: the time-saving benefits combined with that all-important human touch.
You can find out more about LeonidLive here.