About the Cognitive Profile Proof of Concept 

The brain operates through electrochemical signalling in which neurons generate electrical impulses when stimulated by internal or external inputs. These impulses propagate as rhythmic oscillations that can be detected on the scalp using EEG technology. The measurable characteristics of these signals – primarily frequency (how fast neural oscillations occur) and amplitude (the strength or intensity of the signal) – provide insight into levels of cognitive activation within specific brain regions. Raw EEG waveforms can be mathematically transformed into averaged frequency-band and amplitude values, allowing the complex wave diagram to be represented as a simplified bar chart reflecting mean activation for a defined region.

The Five Channel Approach

For simplicity and practical application Competitive Edge Technology (CET) focuses on 5 channels capturing data from the brain regions most relevant to Human Resource management practices. Within the 5-channel classification model, a single composite score is calculated for each region by aggregating these averaged values. Initially, these regional scores were estimated using AI-derived hypothetical cognitive demand values inferred (AI inference engine shown in the diagram below) from large-scale job description analysis across Australian and international occupational sources.

AI has been known to behave in unexpected ways, so we undertook a reasonability check to make sure the values calculated by AI align with expectations based on industry experience.

Validation Study

The validation study that follows replaces hypothetical modelling with real participant EEG testing, comparing measured neural activation against job-based demand estimates to establish empirically validated cognitive demand benchmarks for each occupation.

The 5 Channel Physical Positioning of EEG Headset Electrodes

The BCP Validation Study focuses on five key brain regions, monitored through a 5-channel EEG classification system, selected for their relevance to these cognitive functions. To establish baseline profiles for these regions, AI inference was applied to a rich set of job descriptions sourced from Australian and international occupational datasets. This enabled the identification of hypothetical neural activation patterns and frequency ranges corresponding to each role’s cognitive requirements.

Understanding workforce performance and wellbeing begins with a clear picture of job cognitive demand – the level and type of mental processing required to successfully execute the tasks and responsibilities associated with a role. In neuroscience terms, cognitive demand reflects the activation, coordination, and sustained engagement of specific brain regions, each contributing to functions such as executive reasoning, creative problem solving, attention control, memory integration, and persistence under pressure.

Working with the SENA Network Community

The next phase of the validation study will focus on empirical validation. This will involve working with Australian companies to test a representative set of benchmark roles using EEG headsets, measuring real-time neural activity during job-relevant tasks. The objective is to assess how closely the AI-derived Channel values align with observed neurocognitive patterns, thereby validating the model’s assumptions and outputs.

In the validation study, jobs themselves serve as reference points: by mapping roles within a company to the pre-validated Baseline Cognitive Profiles (BCPs), organizations can evaluate how individual employees’ cognitive capabilities align with job requirements. This comparison helps identify cognitive gaps, supporting targeted development, role alignment, and strategies to enhance both employee performance and wellbeing.

A critical element of this modelling process is the documented AI reasoning stored within CET’s Confluence knowledge base. For each of the 1,500 jobs evaluated, the AI-generated feedback includes the extracted task attributes, identified cognitive demand indicators, weighting logic applied to each channel, and a written rationale explaining how the final channel values were calculated. This audit trail is an invaluable asset. It allows traceability from the published hypothetical channel value back to the specific occupational descriptors and cognitive criteria that informed it. It also enables structured reasonability checks, methodological review, regulatory defensibility, and refinement prior to publication.

Importantly, access to this detailed inference documentation and the associated validation methodology is restricted to CET clients. The Confluence repository represents proprietary intellectual capital that underpins the transparency, auditability, and scientific governance of the BCP framework.

Data Storage and Strategic Use

CET’s BCP Validation Study is underpinned by a purpose-built data architecture designed to treat employee neurodata as highly sensitive, regulated information while still enabling its strategic and ethical use. At the core sits a secure Unified HR Database that separates personally identifiable information from neurocognitive metrics through structured data partitioning, encryption at rest and in transit, and strict role-based access controls. The front-end ecosystem – comprising the Confluence Knowledge Base (for AI rationale, governance traceability and controlled knowledge sharing), the Salesforce transaction and business data repository, and the HR neurotechnology module – operates through permissioned integration rather than unrestricted data exposure.

Custom fields and modular design allow organisations to configure workflows, reporting layers and consent frameworks without altering the protected core dataset. This flexible architecture avoids the risks associated with monolithic centralised systems, enabling clients to maintain privacy, compliance, and auditability while still leveraging validated cognitive insights for workforce planning, capability development and wellbeing initiatives.

International Regulatory Compliance

From a regulatory perspective, the Baseline Cognitive Profile (BCP) framework sits within the evolving landscape of the EU Artificial Intelligence Act, particularly where AI systems are used in employment-related contexts. Under the EU framework, AI systems that assist in recruitment, worker evaluation, or performance assessment may be classified as high-risk due to their potential impact on individuals’ livelihoods. The BCP model aligns with the Act’s objectives in that it emphasizes transparency, human oversight, measurable biological data inputs, validation against real-world testing, and the avoidance of opaque or purely inferential decision-making. Where it differs is in its reliance on direct physiological measurement (EEG-derived neural activation) rather than behavioural profiling or predictive personality inference. While this distinction does not automatically remove it from regulatory scope, it strengthens its defensibility under requirements relating to data quality, explainability, risk management, and fundamental rights safeguards. Accordingly, EU AI compliance should be viewed not as a barrier, but as a structured governance pathway—one that can be addressed through sandbox validation (more details about CET’s generic sandbox will be published shortly), documented human oversight, and rigorous scientific benchmarking rather than as a show-stopping constraint on innovation.

Australian AI Usage Issues

In relation to trade union oversight and workplace health and safety (WHS) scrutiny, the BCP framework should be understood as a protective and preventative tool rather than a surveillance mechanism. The model does not seek to monitor workers in real time, predict behaviour, or replace managerial judgment. Instead, it is designed to identify cognitive demand characteristics at the job level and compare these against voluntarily obtained, controlled EEG baseline measurements under transparent testing conditions.

From a WHS perspective, this aligns with established obligations to identify psychosocial hazards, manage cognitive overload risks, and implement evidence-based controls to protect mental health. Importantly, any deployment of BCP within an organisation would require informed consent, clear purpose limitation, strict data governance protocols, and separation between health-related insights and disciplinary processes. When structured appropriately, the framework can strengthen, rather than undermine, worker protections by providing objective evidence to support safer job design, workload management, and reasonable adjustments. As such, union engagement and WHS compliance should be built into governance architecture from the outset, positioning BCP as a risk mitigation instrument rather than a workforce control mechanism.

Follow Up BCP in Practice – Proof of Concept

A follow up to this article will provide an analysis of the five EEG channels underpinning the BCP. The analyses will examine each channel in detail, illustrating the cognitive demands of the top jobs per channel and extrapolating insights to the broader set of 1,500 roles in the study. By combining EEG-derived individual cognitive profiles with job cognitive demands, the BCP Validation Study aims to provide a scientifically grounded framework for understanding, measuring, and optimizing the cognitive capital of the workforce. The following is an example of what is to follow over the next few weeks:

Channel 1 – Executive / Strategic Cognition

(Prefrontal Cortex – typically Fp1/Fp2, Fp3/Fp4)

Cognitive demands

This channel represents:

• Planning and sequencing of actions
• Logical reasoning and problem-solving
• Decision-making under uncertainty
• Risk assessment and prioritisation
• Cognitive control and inhibition

It is heavily engaged in:

• Policy formulation
• Strategic planning
• Complex judgement calls
• Multi-constraint decision environments

What a high score tells us

A high Channel 1 score indicates the job:

• Requires sustained strategic thinking rather than reactive task execution
• Involves frequent judgement calls with consequences
• Cannot be reduced to procedural automation
• Places high mental load on executive function

Implications:

• Higher fatigue risk under cognitive overload
• Strong suitability for AI decision-support (not decision replacement)
• High alignment with leadership, architecture, governance, and planning roles