CEO confidence in comms is at a record high.

Falling-confidence reports dropped from 13% to 2% in a year. The next test is the AI claim made in a leader's name.

Michael Reaney  ·  18 August 2026  ·  3 min read

The strongest reading yet

Communications leaders have rarely been better trusted by their chief executives. The Confidence Gap study, published on 22 January 2026 by Weber Advisory and Gravity Research from a survey of more than 200 Fortune 1000 communications and corporate-affairs leaders, recorded its strongest year yet: only 2% of respondents report falling confidence from their chief executive, down from 13% a year earlier.

A function that has spent years arguing for its standing appears, on this evidence, to have won the argument. The more useful question is what could reverse the number, and the study itself offers a specific answer.

Where the pressure moves next

Gravity Research's president calls AI a credibility test for leadership, and puts the biggest risk in the widening gap between what companies say AI will deliver and what stakeholders actually experience. The exposure is structural. Generative tools now sit inside everyday drafting, while accountability for what they produce has not moved: a claim that proves wrong lands on the executive whose name is on it, not on the software that drafted it. The better the tools get, the easier it becomes for an unchecked claim to reach publication looking finished.

Two checks that follow

Two questions give a team an early reading of that exposure. The first is whether a named person stands behind every claim published under the chief executive's name, someone who can be asked about a specific sentence rather than a stage in a workflow. The second is whether the answer to "how was this statement produced" would build trust or cost it if a journalist asked.

The second check is also the shape European regulation now expects. Under Article 50 of the EU AI Act, AI-generated text on matters of public interest needs either a disclosure label or a named, accountable human reviewer. A team that can answer the provenance question cleanly tends to be compliant already.