|
IEEE Seattle Section Joint Chapter Secretary Shubham Anandani has spent the past year working on a question that gets harder as AI gets better: how do you know who, or what, to trust?
Anandani first raised the question in December 2025 at IEEE Frontline AI in Seattle, in a talk titled "AI and the Future of Professional Profiles." AI can now write resumes, bios and even entire personas. "We can create more than ever, but we trust less than ever," he told the audience.
In March, a formal version of the work was published and presented at IEEE SoutheastCon 2026, in the paper "Trust-Aware Professional Identity Representation in AI-Mediated Systems." Automated systems screen and rank people at a huge scale, mostly relying on static credentials such as degrees, titles and years of experience. In contrast, human reviewers look for other indicators, including authenticity, and whether they support their peers.
To close that gap, the paper proposes a framework called ACE-U. ACE-U builds on Frances Frei's Trust Triangle of authenticity, credibility and empathy, and adds a fourth dimension, uniqueness, for telling people apart once many candidates clear the trust bar. Each dimension is tied to signals a system could observe. Truth in representation, for example, can be read from whether a person's skills and projects line up over time; and from talks, code and writing that back up their claims. On empathy, Anandani argues that a few detailed endorsements from real collaborators should count more than hundreds of generic ones.
The paper also covers the downsides. Keyword-heavy profiles can look credible to a machine, but less so to a person. Authenticity metrics could penalize people who keep parts of their identity private for safety reasons, and engagement metrics could shortchange caregivers. Anandani calls for configurable weighting and regular fairness audits. The work is conceptual; controlled experiments and field studies are next.
Though the research centers on professional profiles, the underlying question also applies to connected systems, where people, devices and services must decide what to trust. "AI should amplify truth, never invent it," Anandani said in the talk.
Anandani is an IEEE Senior Member. The IEEE Seattle Joint Chapter includes the IEEE Vehicular Technology Society.
|
