Ethical AI Usage

AI never makes the final hiring decision on AETHEON. Our matching, job-description enrichment, and assessment systems surface evidence and flag tensions; the call belongs to people. This is architectural, not aspirational. Tension flagging, per-dimension breakdowns, and editable AI outputs are built in by design so that meaningful humans-in-the-lead is the default state.

AETHEON does not and will never use AI to enhance, augment, embellish, or invent a candidate's skills or experience. Candidates are free to edit and curate their own content, but our AI reports only on the content the candidate has supplied. Skills are extracted, mapped, and structured against our taxonomy, never inflated. The profile a candidate publishes on AETHEON is theirs, drawn from what they have actually done.

We do not ship black boxes. Candidates and employers see the same skill breakdowns, the same alignment data, and the same tension flags. The AETHEON Skills Atlas™ taxonomy and proficiency model provides a consistent framework behind every match, so both sides of the marketplace understand what is being measured and why.

Our AI focuses on the skills people demonstrate, not demographic proxies, name origins, school prestige, or geography.

In our Skills Quests™ validation platform, we actively monitor for behaviors associated with external AI use during an assessment. This includes response timing patterns, copy and paste activity, the use of code or programmatic responses, and other signals consistent with off-platform assistance. We report this activity as data, without assumptions, scores of suspicion, or pass/fail judgments. The candidate and the reviewing party see the same record (summary only for reviewers) and can have an informed conversation about it. We believe respect for candidates means helping them understand the why, not simply labeling the outcome.

AETHEON members own their content and control who sees it. Profiles can be made public, restricted, or fully anonymous to employers inside and / or outside the platform. When an employer is interested in a candidate whose identity is gated, the candidate is approached on their terms and decides whether to release their information. No silent scraping. No back-channel resume passing. Every connection is initiated with consent.

We use AI to expand how a person is seen, not to compress them into a single score. AETHEON.ME profiles surface values, working style, growth trajectory, and lived experience alongside skills. AI translates evidence into structure; it does not flatten the human, quite the opposite.

AI models reflect the data they learn from. We acknowledge this openly in our product disclaimers, design our assessments to reduce common bias vectors (which is why we favor audio over video in AETHEON.ME profiles, reducing appearance and AV-quality disparities), and architect our systems for the bias audits that emerging regulation requires.

We collect what is needed to deliver value, and no more. Users own their data and their profiles. We do not sell candidate data. We process documents and assessment responses solely to deliver the service the user requested. Records a user intentionally saves as part of a profile or job application may be retained to provide that workflow; transient AI prompts and responses are not retained by AETHEON.

AI outputs are estimates that carry uncertainty, and we say so. Every AI-generated skill suggestion, match score, and trajectory projection is presented as a starting point for human judgment, not a verdict. Users can review, edit, override, or remove AI-generated content from their profiles at any time.

We designed our 8-Dimensions of Work Culture DNA to the standards of the EU AI Act, GDPR Article 22, and NYC Local Law 144 even where we are not yet legally required to. Treating these as a floor rather than a ceiling protects the people on our platform and prepares us for the global convergence on responsible AI in hiring that is already underway.