If the only intelligence you’re thinking of is artificial, your compliance and risk programs are missing out on a crucial element, writes Chris Tamdjidi of Awaris, a consulting and training company. Physiological intelligence is more than trusting your gut, it’s about training your intuition to make helpful risk decisions.
The COO of a global logistics company once described his risk radar in disarmingly simple terms: “Give me 15 minutes in any facility, and I can feel what’s wrong.” He wasn’t talking about dashboards or audit reports but something harder to defend in a board paper: a felt sense. A pattern of signals — a supervisor’s tension, a half-second pause before answering, managers avoiding eye contact — that his body forms into a hypothesis before his conscious mind can explain it.
For a risk and compliance audience trained to value documented evidence, this kind of statement sits uncomfortably. Yet, his track record was hard to dismiss. The places that gave him a bad feeling were, often enough, the places where something was actually going wrong.
This is physiological intelligence: the trained capacity to read the signals your body and senses are picking up and to treat them as a legitimate source of information about a complex system.
How it actually works
There is nothing mystical about this. Our conscious, deliberative mind, the part that writes the risk register, is limited in throughput. Most of us struggle to multiply two two-digit numbers in our heads. Meanwhile, non-conscious processing — recognizing a face in a crowd or sensing a shift in tone — runs in parallel at far higher bandwidth. If conscious thought is loose change in your pocket, non-conscious processing is the entire economy.
Emotions and bodily sensations are one of the main ways that the non-conscious system reports back. A flicker of unease on a shop floor is not noise but the output of a pattern-recognition engine drawing on thousands of past cues — micro-expressions, sounds, culture — and flagging a deviation. The COO’s 15 minutes were not magic but a lifetime of pattern recognition, delivered as a feeling.
Crucially, a felt sense is not the same as the truth. It is a hypothesis, not a verdict. But it is a trainable hypothesis generator, and ignoring it is a form of negligence.
Why compliance has always quietly relied on this
Risk professionals do this constantly even if they don’t describe it that way. An auditor asks one more question because something doesn’t sit right. A compliance officer escalates a third-party relationship she can’t fully justify on paper. An investigator senses a witness is holding back. We dress these moments in procedural language, but the initial signal is often physiological.
A 2011 study of Israeli parole judges by Ben-Gurion University and Columbia University professors showed the cost of not listening to these signals. Across 1,100 hearings, favorable rulings dropped sharply, from around 65% at the beginning of sessions, not long after breakfast, to almost none as the early sessions came to a close. The favorable rulings returned to about 65% after breaks, when judges had eaten. The judges would likely have denied that hunger or fatigue were shaping their decisions, but their bodies were signalling depletion. This showed up as a systematic, invisible bias, one that made them more prone to irritation, more likely to dismiss applications and more likely to reject them, holding real consequences for the people in front of them.
The lesson for compliance leaders isn’t just “trust your gut.” It is instead, “listen to and train your gut signals.” The training part is important: Unless you train people to notice and interpret these signals, they will quietly steer decisions anyway and usually in ways nobody is accountable for.
Why Experience Still Matters in an Automated Finance World
AI is reshaping workflows in finance, but the judgment that protects organizations remains deeply human
Read moreDetailsIt is tempting to assume that as AI systems take on more of the analytical heavy lifting in risk and compliance — anomaly detection, transaction monitoring, control testing — the human “felt sense” becomes a quaint relic. Actually, I believe the opposite is true.
AI excels at processing what has been encoded but is blind to what hasn’t. It cannot see that a critical control team is exhausted and close to breaking. It cannot sense when overconfident outputs go unchallenged because no one feels safe to speak. These are the failure modes behind major compliance breakdowns, and they sit in the human layer of human-AI systems.
Physiological intelligence is how we surface the risks that sit between the model and the people using it. The more decisions we delegate to AI, the more critical it becomes that humans can read the felt signals the system cannot generate.
As we are inundated with data, screens and insightful analysis, a quieter risk emerges, visible on any train or plane: No one is looking up anymore. Few are checking what they see, feel or hear. Our physical senses risk degrading as we become entranced by screen-mediated intelligence, increasingly trusting what data and AI tell us without checking our own sensory perception or gut.
What to do
Treat physiological intelligence the way you treat any other control: as something that must be trained, practiced and embedded in habits, not assumed.
Build short body-and-mood checks into risk meetings. Before reviewing AI outputs, ask what people are noticing in themselves and the room. Train compliance teams in mindfulness and interoception, not as well-being perks, but as risk-sensing tools. Make it psychologically safe to say “something doesn’t feel right, but I can’t tell you why,” and create space to investigate rather than override it.
Interestingly, none of this is new. Research on high-reliability organizations by Karl Weick and Kathleen Sutcliffe shows that businesses operating under extreme risk rely on collective mindfulness, attention to frontline operations and deference to the expertise of those closest to the work, including the felt sense that those people bring. Crucially, this depends on environments where people feel safe to speak up when something doesn’t seem right.
The COO’s 15 minutes were not a substitute for data. They were an early-warning system that data alone could not provide. In an AI-driven workplace, that early-warning system is no longer a nice-to-have. It is part of how grounded, balanced, defensible risk judgement actually gets made.
Chris Tamdjidi, co-founder and managing director of Awaris, an international consulting and training company. He works with global organizations on resilience, leadership and the human dimensions of transformation. 
