My weekly digest of five AI+PR updates.
1. When an executive writes on LinkedIn, AI is paying attention.
Profound analyzed 1.4 million citations and found LinkedIn is now the #1 cited domain for professional queries across ChatGPT, Gemini, Google AI Mode, Copilot, and Perplexity. Here’s what that means in practice: when someone asks an AI about a topic that your client has written about, that post can be pulled and cited as a source. Their analysis, their POV, their expertise feeds directly into how AI explains that subject to the world. The content cited most? Original posts and long-form articles from individual creators, not company pages. Reshares account for just 5%. This isn’t a social media play anymore. It’s an AI visibility play.
2. Local TV isn’t what it once was, but that’s not the point.
Local TV viewership is declining but every segment that airs also gets posted to a station’s website, transcribed, and indexed. That digital trail is what AI platforms crawl, cite, and surface long after broadcast. This is the shift from SEO to GEO or Generative Engine Optimization. Where SEO was about keywords and links, GEO is about authority, structure, and alignment with the questions people are actually asking AI.
3. 93% of AI search sessions end without a click. That’s the new normal.
More than half our AI search sessions end without anyone visiting a website. AI overviews actually reduce clicks to top-ranked pages by 58%. That means the way your client is described, evaluated, and recommended is dictated by AI-generated answers, not your own website or messaging. The brands showing up accurately in those answers have one thing in common: a sustained earned media footprint, that is press coverage, analyst mentions, third-party citations, Wikipedia entries, LinkedIn publishing. If your client isn’t visible in those places, AI is either guessing or silent. Both are a problem. And AI visibility isn’t static; one study found a brand can lose a third of its AI presence in just over a month. Quarterly audits are not enough.
4. 55% of companies have no crisis plan for AI-generated misinformation. That’s the next PR emergency.
Signal AI’s 2026 Velocity of Disinformation report found that while 98% of communications professionals see misinformation as a major threat, more than half have no formal plan to address it. False stories travel six times faster than the truth and can reach 100,000 people before a correction reaches 1,000. Deepfake fraud is up 3,000%. Digital deception costs the global economy $78 billion a year. The crisis playbook just got a new chapter. Monitoring coverage and responding to bad press is no longer enough; you need to monitor what AI platforms are saying about your client. A crisis “resolved” two years ago may still be framed as a current risk by an LLM drawing on an old digital trail. The brands that win aren’t the ones with the fastest lawyers. They’re the ones managing their AI footprint before something goes wrong.
5. Meta built an AI Zuckerberg to talk to employees. PRs should be uncomfortable.
Meta has deployed an internal AI chatbot modeled on Mark Zuckerberg — trained on his statements, tone, and strategic thinking so 70,000+ employees can “interact” with a digital CEO. The goal is scale. The problem is trust. As one crisis expert noted, employees expect authenticity from their executives, and “that is not what this is.” An AI trained on past statements can inform and echo but it can’t listen, adapt, or take accountability. Those things are still irreducibly human. The irony: the more companies automate executive voice, the more valuable a genuine, present human communicator becomes.
For PR advice and guidance during this time of AI transformation, don’t hesitate to reach out: Ruth@sarfatycomms.com