AI brands are hiring to reshape narratives around safety

Remember last month’s troubling OpenAI-Hugging Face incident, in which OpenAI agents escaped a test environment and hacked the AI platform Hugging Face?

OpenAI has posted multiple open roles dedicated to communicating to the public its effort to deploy AI responsibly, according to job listings viewed by Marketing Brew. One role, for a strategic narrative and impact lead on OpenAI’s Intelligence and Investigations team, is aimed at helping communicate the company’s safety efforts, according to the job listing, including creating “‘evidence-backed safety impact retrospectives’ that facilitate safety publications, threat reports, and workshops designed for engagement while also not creating “unnecessary additional reporting burdens.”

“Your work will help senior leaders and the public understand the impact of our safety efforts, support thoughtful engagement with external partners and government stakeholders, and make the team’s work more coherent and accessible across the company and beyond,” the job listing, which has since been taken down, read.

Another open role at the company, which is still live, is for a safety transparency editor on the company’s “Trustworthy AI Team” embedded in OpenAI’s Safety Systems division, which works to explain OpenAI’s safety protocols. Job responsibilities include supporting the production of technical materials like system cards and blogs focusing on safety.

“Your core responsibility is to shape and execute how our technical safety work is externally communicated: identifying the narrative thread, exercising judgment about which details matter, determining where additional context, explanation, or supporting evidence is needed, translating complexity without sacrificing precision, and helping external audiences understand both the safety measures we’ve taken and the uncertainties that remain,” according to the job listing.

OpenAI did not respond to a request for comment.

The job listings are part of a broader trend of AI companies looking to bring on people whose jobs entail helping to communicate company safety efforts to the public, as fears regarding AI safety remain heightened and amid several AI-related safety scandals. After OpenAI disclosed the Hugging Face incident, Anthropic reported last month that three separate incidents occurred where agents were able to access the internet to hack other organizations.

Consumers also have their own reservations around AI regulation: 59% of US adults indicated they have “little or no control” over AI use in their personal lives, according to a study conducted by the Pew Research Center in April.

Safety (messages) first: OpenAI’s hiring efforts follow fellow AI companies like Anthropic and Cohere, who have in recent months also listed openings for roles designed to shape a narrative of responsible AI deployment. Cohere recently posted a job listing for a senior manager, global public affairs and policy narratives, tasked with “establish[ing] thoughtful messaging strategies to address sensitive policy inquiries regarding security, open-source models, and copyright,” according to a job listing.

And Anthropic earlier this year posted an open role for a head of GTM narrative with job responsibilities that include communicating the “transformative potential of safe AI” to enterprise organizations, per the listing.

Neither Cohere nor Anthropic responded to a request for comment.

Give it your all: OpenAI and Anthropic have been hard at work placing ads to build awareness of their products. Both have been increasing their share of ad impressions in certain metros, including Philadelphia and Detroit, Marketing Brew previously reported. OpenAI also peppered major cities like Miami and LA with a creator-focused ad campaign in June.

In July, Anthropic appeared in the World Cup broadcast with a 90-second ad that seemed designed to engage with responsible use of AI, although the imagery in the ad, which features a burning house and other apocalyptic imagery, creeped some people out, according to TechCrunch.