PR shifts to real-time fact-checking as AI fakes surge

By mid-2026, communications leaders were describing their work less as message production than as a live verification system for organizations navigating AI-distorted information. In Bizcommunity, Natalia Rosa said that in “a world of information overload and AI-generated everything,” the core task is to “establish what is true, position it clearly, and build trust around it,” while warning that when algorithms read reviews, earned media and third-party mentions to decide what to recommend, reputation and narrative become operational inputs that communications teams must continuously monitor, interpret and manage.

That logic quickly hardened into process: Forbes argued communications leaders must be embedded in risk management before crises strike, with trust monitoring, AI search visibility oversight and prebuilt executive narrative packages ready for misinformation events. Brand Icon Image made the same point at the response level, saying companies should preassign who monitors online conversations, verifies emerging claims and approves statements, a need illustrated by “the controversy involving Erisco Foods and consumer Chioma Okoli in 2023,” where Okoli’s Facebook complaint about Nagiko Tomato Mix rapidly widened into a reputational flashpoint.

The operational shift starts with literacy and oversight: SAS argued that AI value depends on governance frameworks that keep people responsible for goals, guardrails and high-risk decisions, with “human in the lead” rather than ceremonial review. That logic is already being built into enterprise routines; at Voya, AI training was “for our entire workforce – at every level, within every function,” almost its full base of 11,000 employees participated, “In just three months, nearly 100% of employees across titles, functions and backgrounds achieved their literacy certification,” and “98% of our employee base is actively using Copilot with 24 average weekly prompts per user.”

What communicators are building next looks less like a content studio than a control room: Mara Hedgecoth said “There is much less appetite for commentary that appears performative,” while APCO found 72% of Americans respected companies that held to existing commitments under pressure, pushing teams toward decision rules, monitoring and pre-approvals. Collins cited the Sonos subreddit after a “messy app update” in 2024, noting that the value wasn’t “bland ‘we’re listening’ statements,” but that Sonos reps could “answer with real operational information,” showing how real-time verification and response depend on cross-functional access and crisis playbooks, not just publishing speed. That same governance turn is being formalized by ICCO, which on 16 September 2026 launched an “industry-first Mis/Disinformation Training & Certification Program” to help professionals “identify, counter, and prevent” falsehoods; the course runs “from October 15 to December 15, 2026,” spans “six core modules plus final test,” and ends with “Final Test and Certification – Earning industry-recognised credential.”

The damage from AI-generated deception is no longer hypothetical: Businessday NG reported that a 2026 Surfshark analysis estimated global deepfake fraud losses at $3.7 billion through June 2026, with social media accounting for $1.73 billion, or 47 percent, of documented losses. SEC Newgate UK underscored why those losses can scale so quickly, warning that the rapid evolution of public AI models has made disinformation easier to create and spread at scale, while 28% of users of the technology believe it “always produces factually correct information,” making false outputs unusually potent.

The reputational fallout is equally concrete in high-profile impersonation cases that hijack institutional credibility: in April 2024, scammers circulated a doctored Channels Television video making it appear Aliko Dangote was promoting an AI-powered investment platform, and both Dangote Group and Channels Television distanced themselves from the scheme. Businessday NG also noted that in May 2025 NNPC Limited warned Nigerians about an AI-generated video featuring a cloned voice of Group Chief Executive Officer Bayo Bashir Ojulari, and in May 2026 the Presidency raised concerns about a deepfake involving President Bola Ahmed Tinubu with a false religious narrative that could inflame tensions.