Why It’s Important to Know AI Limitations #
The platform uses artificial intelligence models to generate conversations, interpret communication signals, guide the avatar, and produce training reports.
While responses may appear natural, coherent, and realistic, it’s important to remember that AI can make mistakes.
AI-generated responses may be:
- plausible but not entirely correct;
- consistent in tone but inaccurate in content;
- convincingly formulated but not suitable for use as an official source;
- influenced by the simulation context, Avatar Persona, and training objectives.
For this reason, AI-generated content should be read as part of a simulation and learning experience, not as absolute truth.
What It Means That AI Can Make Mistakes #
AI does not “know” in the human sense of the term.
It generates responses based on the context received, data available in the simulation, and the language models it has been configured with.
This means it can produce content that seems credible but may contain:
- inaccuracies;
- omissions;
- imperfect interpretations;
- generalizations;
- context errors;
- outdated information;
- plausible but unverifiable statements;
- logical connections that are not always correct.
This phenomenon is often called AI hallucination: a response that appears plausible but is not necessarily true or 100% reliable.
The Avatar Is Not an Official Source #
The avatar is designed to simulate a realistic interlocutor and generate a conversation useful for training.
It should not be considered an official source for operational, legal, medical, financial, compliance, HR, or organizational decisions.
During a simulation, the avatar can:
- play a role;
- express doubts, emotions, resistance, or objections;
- react to the Player’s words;
- create a realistic conversational challenge;
- provide information consistent with the scenario.
However, what the avatar says must be interpreted within the context of the simulation.
If relevant technical, regulatory, clinical, legal, or business content emerges, it must be verified with official sources, internal policies, or qualified personnel.
Reports Are Training Tools, Not Absolute Judgments #
Reports generated by the platform should also be read as learning support tools.
The report analyzes conversational signals, Communication Skills, Observation Grid phases, Readiness, and key moments in the conversation.
However, the report should not be interpreted as:
- a diagnosis;
- an absolute certification of competence;
- a definitive assessment of the person;
- an automatic HR judgment;
- a disciplinary decision;
- an infallible measurement of performance.
The report helps understand what emerged in a specific simulation and which behaviors can be strengthened in subsequent attempts.
Difference Between Plausibility and Truth #
An effective simulation must be plausible.
This means the conversation must seem realistic, natural, and consistent with the scenario.
But plausible does not necessarily mean 100% true.
For example, in a simulation:
- the avatar can represent a patient, manager, customer, or colleague;
- the scenario can reproduce a realistic situation;
- the conversation can seem very similar to a real case;
- the report can highlight useful signals.
However, all of this remains part of a simulated environment.
The simulation serves to train behaviors, not to replace professional assessments, business decisions, or official sources.
When to Verify Content #
It is always advisable to verify AI-generated content when it concerns:
- regulations;
- company procedures;
- internal policies;
- compliance risks;
- medical or health aspects;
- legal aspects;
- HR decisions;
- workplace safety;
- economic or financial data;
- sensitive operational information;
- official communications to customers, patients, employees, or stakeholders.
In these cases, AI can help simulate a conversation or propose reasoning, but the final decision must always be based on validated sources and human responsibility.
User Role and Human Oversight #
The platform is designed to support learning, not to eliminate the role of human oversight.
Tenant Admins, Providers, trainers, managers, or company representatives should consider AI as a support tool.
Human oversight remains important to:
- validate scenarios;
- control simulation quality;
- verify the accuracy of sensitive content;
- correctly interpret reports;
- contextualize results;
- prevent misuse of generated data.
Correct Use of AI on the Platform #
AI should be used to:
- create realistic conversational experiences;
- train communication skills;
- simulate interlocutors with stable behavioral patterns;
- help the Player reflect on their own communication;
- generate training feedback;
- identify areas for improvement;
- support coaching and practice pathways.
It should not be used as the sole source to:
- decide whether a person is competent or not competent;
- make disciplinary decisions;
- formulate diagnoses;
- provide legal, medical, or financial advice;
- replace policies, procedures, or official documents;
- automatically certify a person’s suitability for a role.
Why This Transparency Is Important #
Declaring AI limitations does not reduce the platform’s value.
On the contrary, it makes its use more correct, safe, and professional.
The platform is valuable precisely because it allows the creation of realistic, observable, and repeatable simulated environments.
But the training value comes from conscious use of the tool:
- AI generates the experience;
- the Player trains;
- the report helps reflection;
- the organization interprets data in the correct context;
- human oversight maintains final control.
Final result #
AI can generate very realistic responses, but should not be considered infallible.
Avatar responses, interpretations, and reports should be read as simulation and learning tools.
In the presence of sensitive, technical, or decision-making content, it is always necessary to verify information with official sources, qualified experts, or internal procedures.
The platform’s goal is not to replace human judgment, but to support training, reflection, and improvement of communication skills.
