Generative AI is making individual pieces of a synthetic identity easier to manufacture. SEON is responding by expanding its Signal Intelligence platform from more than 900 to more than 1,100 proprietary, directly sourced data points spanning address data, session behaviour, phone and carrier information, device intelligence and digital footprints.
The headline signal count should not be treated as a universal performance benchmark. SEON itself published a separate article noting that the fraud industry lacks a shared definition, independent benchmark or standard audit for signal counts. The more useful story is the underlying defensive model.
SEON argues that a convincing image, document or voice sample can increasingly be generated on demand, while building a coherent history across multiple independent dimensions is harder at scale. Its thesis is that reused devices, infrastructure, carrier data, session behaviour and digital-footprint inconsistencies can expose fraud rings even where individual identity artefacts look plausible.
That is a vendor position, not an independent proof that this particular approach outperforms competing fraud systems. But it captures an important direction for financial crime defence: the question is shifting from whether one artefact looks real to whether the whole identity and its surrounding behaviour make sense.
In financial services, that matters for synthetic identity, account opening, mule activity, remote access and agent-mediated transactions. The contest is moving from fake versus real content toward fabricated versus consistent context.
