Fama Debuts Policy-Based AI Screening With New 'Lens' Product

Frequently Asked Questions

What is Fama Lens and how does it differ from traditional flags-and-keywords screening?

Fama Lens is the company's first policy-based screening product, announced for the first time on this episode. Instead of smashing Fama's flags and a customer's policy together, Lens deconstructs the customer's governance policy into a multi-agent consensus architecture where agents must agree before adding or removing items from a report. It does not score candidates or issue a yes/no verdict.

How does Fama Lens handle context and regional nuance?

Mones describes a global customer scenario where Lens recognized that 'KKK' written in a Portuguese-language post in Brazil means laughter rather than a reference to the hate group. In another test, Lens flagged an ice cream cake recipe for an influencer-vetting client because the recipe contained rum, and the client, which markets to kids, confirmed they wanted that flagged.

What happens if an employer does not already have a screening policy?

Fama proactively resources an onboarding team that runs an adjudication policy-building session with the customer. That conversation is recorded and transcribed, an agent drafts a policy from it, and the draft goes back to the customer to be reviewed, stamped, and routed through legal before anything is screened against it.

Why does Mones say policy-based screening is actually more compliant than older approaches?

Because every agent decision is traceable back to a specific policy element, and end users never see content that is not mapped to that policy. Mones argues this aligns with principles of least intrusiveness and minimized invasiveness, and that lawyers respond well because the system becomes rules-based, which matches how legal teams are trained to think.

Where in the hiring funnel does Fama recommend running social media screening?

Historically customers have used Fama at the 2-to-3-candidate shortlist stage, at conditional offer, or on annual or biannual rescreening. Mones says Fama avoids real-time monitoring as too Big Brother. His hypothesis is that Lens, because it locks in to an employer's definition of great, will move screening earlier, onto a shortlist of four or five candidates per role.

What did Mones say about identity verification and North Korean applicant fraud?

Mones calls verification the fastest-growing part of background screening over the trailing 18 months, driven by the North Korean fraud problem. He references State Department guidance to employers and screeners, and notes payroll companies are sometimes the ones catching fraud during direct deposit setup, by which point the employer has already exposed roadmaps and confidential information through interviews.

What governance and safety controls ship with Fama Lens?

Mones says Fama placed deterministic rule layers on top of the agentic layer, added a kill switch, and monitors for active model drift. As part of the initial pitch, Fama sends customers the full governance policy for the software itself, along with the customer's drafted policy and logs showing how the AI agents executed against it.