8 Compliance Checks Buy-Side Firms Run Before an Expert Network Call

Every expert call a hedge fund or long-only manager books carries a compliance tail. Before the analyst dials in, the expert network has run the consultant through a layered screening sequence, and the buy-side firm's own compliance team has typically added a second pass on top. This piece walks through the eight checks that show up in a standard pre-call workflow, what each one is trying to catch, and how the major networks operationalize them. It is written for compliance officers, research COOs, and analysts onboarding to expert networks for the first time.

The checks below are layered, not sequential alternatives. An expert can clear gate one and fail gate seven. Audit trails for each step are increasingly a line item in buy-side RFPs, which is part of why every covered network now publishes a public-facing compliance page.

1. Employer screening against a restricted list

The first gate is the simplest to describe and the hardest to get wrong. The buy-side client uploads a list of tickers and entity names the desk does not want to talk to anyone currently employed by. The network's platform blocks any consultant whose current employer matches.

This is the baseline check that every covered network publishes as a core compliance feature, including GLG, Guidepoint, AlphaSights, and Third Bridge. The restricted list is usually maintained by the client's compliance team and refreshed on a rolling basis as positions open and close, which means the screening has to run against the list as of the call date, not as of the project kickoff.

2. Recent-employment cooling-off period

The second gate catches the consultant who left a covered company recently enough that their knowledge is still close to current. The standard cooling-off period is six or twelve months from the date the expert left a restricted issuer. Some buy-side firms extend the window to twenty-four months for companies under active diligence or for sectors where product cycles are long.

The logic is straightforward. A former director of supply chain who left a covered semiconductor name three months ago is not a safe source for a call about that company's next quarter, even if the topic is framed as industry-wide. The cooling-off window is a blunt instrument, but it is the cleanest way to enforce the distinction between durable industry knowledge and recent inside knowledge.

3. Public-company officer and insider screening

The third gate flags or excludes Section 16 officers, board directors, and named executive officers of public companies. This maps directly to SEC Regulation FD exposure, the rule against selective disclosure of material non-public information by public-company insiders that the SEC finalized in 2000 and published in the final rule release.

A sitting CFO of a covered issuer is, in practical terms, a walking MNPI risk. Most networks will not accept the consultant for the topic at all. A sitting independent director may be accepted for an adjacent industry topic, but never for a call about the company on whose board they sit. This screening is run against current filings, and the network's compliance team typically owns the check rather than leaving it to the analyst.

4. Confidentiality and NDA scope check

The fourth gate is an attestation. The expert confirms, in writing, that they are not bound by a confidentiality agreement or NDA covering the specific topic the buy-side client wants to discuss. The important detail is that the attestation has moved from blanket to topic-specific. Networks including Coleman and Dialectica require the consultant to attest against the actual question set or topic brief, not just to a generic "I will not share confidential information" clause.

The shift to topic-specific attestation matters because most former employees are bound by some surviving NDA. The relevant question is whether the scope of that NDA touches the scope of the call. A former engineer can talk about general process knowledge without breaching an NDA that covered a specific unreleased product; the topic-level attestation is how the network documents that distinction.

5. Government and policy-official screening

The fifth gate is a separate workflow for current government employees, recently departed regulators, and registered lobbyists. This screening was shaped by the political-intelligence scrutiny that followed the 2012 STOCK Act, which extended insider-trading prohibitions to members of Congress and federal employees and put the political-intelligence industry under closer congressional review.

In practice, the workflow looks different from corporate screening. A current federal employee is typically declined outright. A recently departed regulator faces a cooling-off period that can be longer than the corporate equivalent, sometimes running to a full year or more, depending on the agency and the topic. Registered lobbyists are accepted in some networks for policy-color calls but flagged so the buy-side analyst knows what they are getting.

6. Custom client-specific exclusions

The sixth gate is where the buy-side firm's own compliance posture overlays the network's. The client uploads a custom list of counterparties, portfolio companies, deal-related entities, and any names tagged as MNPI-restricted on the firm's internal watch list. The expert network platform enforces these on top of its own restricted list.

The two lists are additive. The network's list catches the consultant who works at a public issuer the client trades. The client's list catches the consultant who works at a private company the client owns a stake in, or at a target in an unannounced deal, or at a counterparty the desk is in litigation with. Neither list is a substitute for the other, and the audit trail typically records both checks separately so the compliance team can show which list caught a given exclusion.

7. Topic and question pre-screening

The seventh gate reviews the analyst's question set itself. The questions are checked against a "do not ask" template covering material non-public financials, unannounced M&A, unannounced commercial agreements, and clinical trial outcomes ahead of public readout. Networks including Capvision and ProSapient log this review as a compliance artifact attached to the call record.

This gate exists because the first six checks address who is on the call, not what gets asked. A perfectly cleared consultant can still be walked into an MNPI question by an analyst who does not know the line. The pre-screen is partly a coaching function for newer analysts and partly a documented record that the question set was reviewed before the call happened. For the compliance team, the artifact is what allows them to defend the call later if a position taken after the call is questioned.

8. Recording, transcript, and AI-ingestion consent

The eighth gate is the newest. Experts now consent specifically to the call being recorded, transcribed, and ingested into downstream AI workflows, including LLM-based search, retrieval-augmented generation over the firm's transcript library, and Model Context Protocol pipelines that expose transcripts to internal research agents.

This consent is distinct from the older recording-permission clause. A consultant who agreed to be recorded in 2019 did not agree to have that recording embedded, chunked, and surfaced by an internal LLM in 2026. Networks have responded by adding a specific consent clause covering downstream AI use, and by tightening the redaction layer that strips identifying detail before transcripts move into the model layer. The shape of this gate is still settling, and it is the one most likely to appear as a new line item in the next round of buy-side RFPs.

What the layered model is actually doing

The eight checks above are not a funnel, they are a mesh. Each gate is designed to catch a different category of risk: employment, recency, insider status, contractual obligation, public-official status, client-specific conflict, question scope, and downstream data use. A consultant who passes one is not closer to passing the next; the gates are independent tests against independent risks.