{"componentChunkName":"component---src-templates-insight-jsx","path":"/insight/agentic-ai-what-it-means-for-communicators-and-bra","result":{"data":{"prismic":{"insightByUID":{"title":[{"type":"heading1","text":"Agentic AI: What It Means for Communicators and Brands","spans":[],"direction":"ltr"}],"category":{"__typename":"PRISMIC_Category","name":"In-house Experts"},"author":{"__typename":"PRISMIC_Person","name":"Andrew Zimmer"},"feature_image":{"dimensions":{"width":7815,"height":5351},"alt":null,"copyright":null,"url":"https://images.prismic.io/marino-pr/E058bV9v5PYvVxA5_iStock-2191084357.jpg?auto=format,compress"},"related_case_study":null,"body":[{"__typename":"PRISMIC_InsightBodyText","type":"text","primary":{"content":[{"type":"paragraph","text":"This week, OpenAI launched ChatGPT Work: a version of ChatGPT that connects to Slack, Microsoft Teams, Google Drive, SharePoint, and Salesforce and can carry a multi-step project from start to finish. It competes directly with Anthropic’s Claude Cowork and Microsoft Copilot. OpenAI also released GPT-Live, a voice model that listens and talks at the same time — you can interrupt it mid-sentence, the way you would a person — and a new family of models called GPT-5.6. Anthropic released something different: a dashboard that shows you how you actually use Claude, with optional break reminders and quiet hours.","spans":[{"start":18,"end":39,"type":"hyperlink","data":{"link_type":"Web","url":"https://siliconangle.com/2026/07/09/openai-debuts-chatgpt-work-agentic-tool-automating-business-workflows/","target":"_blank"}},{"start":431,"end":468,"type":"hyperlink","data":{"link_type":"Web","url":"https://www.infoworld.com/article/4195478/openai-launches-chatgpt-work-as-it-broadens-gpt-5-6-rollout.html","target":"_blank"}},{"start":510,"end":564,"type":"hyperlink","data":{"link_type":"Web","url":"https://www.anthropic.com/news/reflect-with-claude","target":"_blank"}}],"direction":"ltr"},{"type":"paragraph","text":"The headlines will say AI can now work “with minimal supervision.” We think that’s the wrong way to understand what just happened — and the wrong thing to want. The real shift is that one person can now run ten workstreams at once instead of two. The question isn’t whether AI can work unsupervised. It’s who supervises it — because someone has to, and that someone is where the value lives.","spans":[],"direction":"ltr"},{"type":"heading3","text":"What Agentic AI Is — and What People Actually Want From It","spans":[{"start":0,"end":58,"type":"strong"}],"direction":"ltr"},{"type":"paragraph","text":"“Agentic AI” is the industry term for AI that completes tasks instead of just answering questions. You give it a goal — build a media list, pull together a coverage report — and it breaks the goal into steps, uses the tools it’s connected to, and comes back with finished work. If that sounds familiar, it should: back in 2023, we took an early look at ChatGPT plugins — small add-ons that let ChatGPT check Zillow listings or read a link — and asked whether they’d amount to anything. They did. Those add-ons were the rough first version of today’s AI agents.","spans":[{"start":336,"end":368,"type":"hyperlink","data":{"link_type":"Web","url":"https://marinopr.com/insight/what-are-chatgpt-plugins/","target":"_blank"}}],"direction":"ltr"},{"type":"paragraph","text":"Here’s what the “minimal supervision” framing gets wrong: nobody is asking to cede control of their work to a machine. Ask the people actually using these tools what they want, and it isn’t an employee they never check on — it’s the ability to do a ton of things at once, with someone they trust making sure it’s all good. Agents still make mistakes and still state wrong things confidently. That’s not a flaw in the vision; it’s the design constraint. The winning setup isn’t autonomy, it’s leverage: agents run in parallel — drafting, monitoring, researching — while a person who knows what good looks like reviews, corrects, and ships. That supervisor can sit inside your team or at a partner you trust. Either way, the new org chart has a layer that didn’t exist two years ago, and “agent supervisor” job listings are already showing up to prove it.","spans":[],"direction":"ltr"},{"type":"paragraph","text":"This is where we think the agency model is headed, and where Marino is investing: not just advising clients on AI, but designing the workflows — the operating system for a brand’s communications — and running them, so clients get the output of ten workstreams with the accountability of one trusted operator.","spans":[],"direction":"ltr"},{"type":"heading3","text":"Voice AI Is About to Change How People Find Brands","spans":[{"start":0,"end":50,"type":"strong"}],"direction":"ltr"},{"type":"paragraph","text":"GPT-Live matters because of what it does to behavior. When talking to an AI feels like talking to a person, more people will ask it things out loud — including things like “what’s a good restaurant near the office” or “which PR firm should we talk to.” A spoken answer is short. It names one or two options, not ten links. There’s nothing to scroll past and no second page. If the AI doesn’t mention you, you were never in the running — and the person asking will never know you existed.","spans":[],"direction":"ltr"},{"type":"heading3","text":"How to Read the Model Wars","spans":[{"start":0,"end":26,"type":"strong"}],"direction":"ltr"},{"type":"paragraph","text":"GPT-5.6 launched with big claims: faster than Anthropic’s Fable 5, 54% fewer tokens (roughly, less computing work) on coding tasks, about a third the cost. Meanwhile, an independent ranking — the Artificial Analysis Intelligence Index — still puts it second to Fable 5. Both things can be true, because companies pick the measurements where they win. We saw the same pattern when we compared ChatGPT and Bard in 2023, and the advice from that piece still holds: the rankings change fast, the tendency to invent facts never fully goes away, and the right tool is the one that fits how your team actually works. Don’t switch because of a launch video. Test on your own work.","spans":[{"start":392,"end":408,"type":"hyperlink","data":{"link_type":"Web","url":"https://marinopr.com/insight/whats-the-difference-between-chatgpt-and-bard/","target":"_blank"}}],"direction":"ltr"},{"type":"paragraph","text":"There’s a second lesson in how these companies fight: every launch comes with a livestream, head-to-head numbers, and a counter-claim within hours. Any brand that communicates about technology should notice what works here — lead with the numbers where you win, expect the other side’s response, and remember that audiences trust independent scorekeepers more than they trust you.","spans":[],"direction":"ltr"},{"type":"heading3","text":"AI Visibility: The Question Every Brand Should Ask","spans":[{"start":0,"end":50,"type":"strong"}],"direction":"ltr"},{"type":"paragraph","text":"“AI visibility” means one thing: when someone asks ChatGPT, Gemini, Perplexity, or Claude a question about your category — “best PR firms in New York,” “top tenant-experience platforms” — does the answer mention you? More buying research now starts with that question instead of a Google search, so the answer matters the way search rankings used to.","spans":[],"direction":"ltr"},{"type":"paragraph","text":"The work of improving it is called “generative engine optimization” (GEO) — you’ll also see “answer engine optimization” (AEO); they mean the same thing. Strip away the name and it isn’t a trick. AI models mostly repeat what credible sources say about you. So the way in is the way it has always been: get covered by publications the AI trusts, publish content that plainly says what you do and for whom, and keep your description of yourself consistent everywhere it appears. That’s the same job as building a content engine — what’s new is that you can now measure the result. Tools exist that track how often each AI platform mentions your brand, and against which competitors. This is exactly the kind of always-on workflow that agents are built for: monitoring, measuring, and flagging changes across platforms continuously, with a person reviewing what matters.","spans":[{"start":500,"end":525,"type":"hyperlink","data":{"link_type":"Web","url":"https://marinopr.com/insight/your-marketing-needs-its-content-engine","target":"_blank"}}],"direction":"ltr"},{"type":"heading3","text":"Your AI Use Will Be Measured — Decide What the Numbers Say","spans":[{"start":0,"end":58,"type":"strong"}],"direction":"ltr"},{"type":"paragraph","text":"Anthropic’s new dashboard looks like a wellness feature — break reminders, quiet hours. The bigger point is that AI use is now something you can see and count. Individuals can see it today; organizations will want the same view soon — which tasks went to an agent, where AI touched client work, and who reviewed it before it shipped. For agencies, this is the opportunity, not the threat. Clients are already asking how AI is used on their accounts. The answer they want isn’t “we don’t use it” or “it runs on its own” — it’s “here’s everything running at once, and here’s the person accountable for each piece.” Being able to show that turns AI from something clients worry about into a reason they hire you.","spans":[],"direction":"ltr"},{"type":"heading3","text":"What Communications Teams Should Do Now","spans":[{"start":0,"end":39,"type":"strong"}],"direction":"ltr"},{"type":"list-item","text":"Ask the AI platforms about yourself. Ask ChatGPT, Gemini, Perplexity, and Claude what they say about your brand and your executives, and who they recommend in your category. Their answers are your new first impression, and your starting point for GEO.","spans":[],"direction":"ltr"},{"type":"list-item","text":"Put agents on the work you can’t staff. Coverage tracking, transcript summaries, first-draft research — tasks you’d do more of if you had more hands. Measure the gain in volume, not just speed.","spans":[],"direction":"ltr"},{"type":"list-item","text":"Build the supervision layer as you scale the agents. Every workflow needs a named person who reviews the output and owns the result. Train for knowing when output is wrong, thin, or made up — that judgment is the new core skill.","spans":[],"direction":"ltr"},{"type":"list-item","text":"Write for the spoken answer. Structure your content so that when an AI reads it aloud, your name and what you do come through clearly.","spans":[],"direction":"ltr"},{"type":"list-item","text":"Write your AI policy before a client or reporter asks for it. The strongest policy isn’t “we limit AI” — it’s “here’s what runs, and here’s who checks it.”","spans":[],"direction":"ltr"},{"type":"list-item","text":"Treat model launches like breaking news. The AI cycle moves in hours, not days. If your clients or executives have a stake in it, be ready to comment the same day.","spans":[],"direction":"ltr"},{"type":"heading3","text":"Beyond the Comms Team: What Brands Should Do","spans":[{"start":0,"end":44,"type":"strong"}],"direction":"ltr"},{"type":"paragraph","text":"AI agents don’t just change how your team works — they change how customers, donors, tenants, and buyers find you, because more of that finding is done by AI on their behalf. The rule we laid out in our intro to integrated strategy applies here too: no more random acts of marketing. Every sector faces the same question — when an AI researches your category, do you show up? Here’s what that looks like in four industries we work in every day. None of it is one-time work; all of it runs best as a standing workflow with an owner.","spans":[{"start":199,"end":231,"type":"hyperlink","data":{"link_type":"Web","url":"https://marinopr.com/insight/an-intro-to-integrated-strategy","target":"_blank"}}],"direction":"ltr"},{"type":"heading4","text":"Food & Beverage","spans":[{"start":0,"end":15,"type":"strong"}],"direction":"ltr"},{"type":"paragraph","text":"People are starting to ask AI, out loud, where to eat and what to buy — “best rooftop restaurants in Brooklyn,” “clean-label snack brands.” The AI builds its answer from what it can read: your website, your maps listing, review sites, and press coverage. So the to-do list is concrete. Keep menus, hours, and locations current and identical everywhere they appear. Tell your sourcing story in plain text on your own site, not just in a PDF or an image. Treat reviews and local press as double-duty: they persuade the humans who read them and inform the AI that answers the next question. And as AI starts handling reservations and orders directly, anything broken in that path — a stale menu link, an unlisted location — costs you the sale. This is exactly the kind of many-small-things-at-once upkeep that an agent-powered workflow handles well, with someone checking the output weekly.","spans":[],"direction":"ltr"},{"type":"heading4","text":"Non-Profit","spans":[{"start":0,"end":10,"type":"strong"}],"direction":"ltr"},{"type":"paragraph","text":"Donors and grant-makers now ask AI whom to trust: “most effective housing non-profits in New York,” “where will my donation go furthest.” AI answers lean on public proof — watchdog ratings, annual reports, financial disclosures, news coverage. If those are outdated or inconsistent with how you describe yourself, the AI’s summary of you will be too. Keep them current. Publish your results as plain, specific numbers on your own site — people served, dollars deployed, outcomes — so an AI quoting you gets it right. The organizations that are easy for a machine to summarize accurately are the ones that get recommended.","spans":[],"direction":"ltr"},{"type":"heading4","text":"Real Estate","spans":[{"start":0,"end":11,"type":"strong"}],"direction":"ltr"},{"type":"paragraph","text":"One of the first ChatGPT plugins we wrote about in 2023 was Zillow’s — ask for listings by budget and bedrooms, get answers in the chat. That was the preview. Now AI agents shortlist neighborhoods, buildings, and brokers before a prospect ever makes a call. What to do: keep listings, availability, and project details current and consistent across every site an AI might read, and get your executives quoted in the trade and business press, because those quotes are what AI cites when someone asks who knows this market. Video still works — we covered how in using video to support your real estate marketing — but add transcripts and written summaries, because AI reads text, not footage. Earned media compounds here too: coverage and commentary, including the housing podcasts we track, are exactly the sources AI answers draw from.","spans":[{"start":560,"end":609,"type":"hyperlink","data":{"link_type":"Web","url":"https://marinopr.com/insight/using-video-to-support-your-real-estate-marketing-efforts/","target":"_blank"}},{"start":759,"end":788,"type":"hyperlink","data":{"link_type":"Web","url":"https://marinopr.com/insight/top-housing-podcasts-to-listen-to-in-2025/","target":"_blank"}}],"direction":"ltr"},{"type":"heading4","text":"PropTech","spans":[{"start":0,"end":8,"type":"strong"}],"direction":"ltr"},{"type":"paragraph","text":"PropTech buying now often starts with an AI agent doing the vendor research: “best tenant-experience platforms,” “leasing automation that works with Yardi.” The agent compares what it can find. If the clearest write-up of your category lives on a competitor’s blog, the agent recommends the competitor. So publish your own comparison content, and make it honest — what you do, what you don’t, which integrations actually exist. Announce partnerships and integrations in specific terms, because AI favors recent, concrete facts over marketing language. And get founders and product leads quoted in trade press: when an AI has to pick whom to believe, third-party sources outrank your own website.","spans":[],"direction":"ltr"},{"type":"heading3","text":"FAQ: Agentic AI for Communications Teams and Brands","spans":[{"start":0,"end":51,"type":"strong"}],"direction":"ltr"},{"type":"heading4","text":"What is agentic AI?","spans":[{"start":0,"end":19,"type":"strong"}],"direction":"ltr"},{"type":"paragraph","text":"Agentic AI is AI that completes multi-step tasks — researching, drafting, and executing across connected tools — instead of answering one question at a time. ChatGPT Work, Claude Cowork, and Microsoft Copilot are the leading examples.","spans":[],"direction":"ltr"},{"type":"heading4","text":"How is agentic AI different from a chatbot?","spans":[{"start":0,"end":43,"type":"strong"}],"direction":"ltr"},{"type":"paragraph","text":"A chatbot answers; an AI agent acts. Agents connect to tools like Slack, Teams, Drive, and Salesforce and can carry a project from start to finish, delivering finished work rather than suggestions. The work still needs review — the point is that one person can now oversee many projects at once.","spans":[],"direction":"ltr"},{"type":"heading4","text":"Do AI agents work unsupervised?","spans":[{"start":0,"end":31,"type":"strong"}],"direction":"ltr"},{"type":"paragraph","text":"They can run without check-ins, but that’s not how they create value. Agents make mistakes and state wrong things confidently, so every serious deployment pairs them with a person who reviews the output and owns the result. The gain isn’t removing people — it’s letting one person direct the volume of work that used to take a team.","spans":[],"direction":"ltr"},{"type":"heading4","text":"What does agentic AI mean for public relations?","spans":[{"start":0,"end":47,"type":"strong"}],"direction":"ltr"},{"type":"paragraph","text":"Two things. PR teams can run far more workstreams at once — monitoring, research, drafting — with people supervising rather than producing every piece by hand. And brands face a new audience: the AI agents that research and recommend companies. You need a plan for using agents well, and a plan for being mentioned by them.","spans":[],"direction":"ltr"},{"type":"heading4","text":"How do brands show up in AI-generated answers?","spans":[{"start":0,"end":46,"type":"strong"}],"direction":"ltr"},{"type":"paragraph","text":"AI models repeat what credible sources say. So: earn coverage in publications AI trusts, publish clear content that says what you do and for whom, and keep your brand description consistent everywhere. This practice is called generative engine optimization (GEO), and tools now exist to measure how often ChatGPT, Gemini, Perplexity, and Claude mention you.","spans":[],"direction":"ltr"},{"type":"heading3","text":"The Bottom Line","spans":[{"start":0,"end":15,"type":"strong"}],"direction":"ltr"},{"type":"paragraph","text":"The race to build your AI co-worker is crowded, and it won’t be won on speed and price alone. It will be won on trust — whose agents people rely on, which brands those agents mention, and who you trust to make sure the work is actually good. That last part doesn’t come from a model. It comes from people who know what good looks like, running systems that let them oversee ten times the work. That’s where communications teams — and the partners who build and operate these workflows with them — earn their place in what comes next. For more on where AI and digital marketing are heading, subscribe to Marino’s AI and Digital Marketing Newsletter.","spans":[{"start":590,"end":647,"type":"hyperlink","data":{"link_type":"Web","url":"https://marinopr.com/insight/introducing-marinos-ai-and-digital-marketing-newsletter/","target":"_blank"}}],"direction":"ltr"}],"alignment":"Left"}},{"__typename":"PRISMIC_InsightBodyCustom_html","type":"custom_html","primary":{"content":[{"type":"preformatted","text":"<script charset=\"utf-8\" type=\"text/javascript\" src=\"//js.hsforms.net/forms/embed/v2.js\"></script>\n<script>\n  hbspt.forms.create({\n    portalId: \"46082381\",\n    formId: \"a01501c4-d8ad-417b-bead-b37a626373a7\"\n  });\n</script>","spans":[],"direction":"ltr"}]}}],"related_insights":[{"insight":null}],"get_in_touch_heading":null,"social_title":"Agentic AI: What It Means for Communicators and Brands","social_description":null,"social_image":{"dimensions":{"width":1200,"height":630},"alt":null,"copyright":null,"url":"https://images.prismic.io/marino-pr/E058bV9v5PYvVxA5_iStock-2191084357.jpg?auto=format,compress&rect=0,624,7815,4103&w=1200&h=630"}},"allSite_settingss":{"edges":[{"node":{"get_in_touch_heading":"Everything starts with a conversation."}}]}}},"pageContext":{"rootQuery":"\n  query InsightQuery($uid: String!) {\n    prismic {\n      insightByUID: insight(uid: $uid, lang: \"en-us\") {\n        title\n        category {\n          ... on PRISMIC_Category {\n            name\n          }\n        }\n        author {\n          ... on PRISMIC_Person {\n            name\n          }\n        }\n        feature_image\n        related_case_study {\n          ... on PRISMIC_Case_study {\n            client_name\n            tagline\n            feature_image\n            area {\n              ... on PRISMIC_Area {\n                name\n              }\n            }\n            _meta {\n              uid\n            }\n          }\n        }\n        body {\n          ... on PRISMIC_InsightBodyText {\n            type\n            primary {\n              content\n              alignment\n            }\n          }\n          ... on PRISMIC_InsightBodyCustom_html {\n            type\n            primary {\n              content\n            }\n          }\n          ... on PRISMIC_InsightBodyQuote {\n            type\n            primary {\n              quote {\n                ... on PRISMIC_Quote {\n                  content\n                  name\n                  title\n                }\n              }\n            }\n          }\n          ... on PRISMIC_InsightBodyContent_block {\n            type\n            primary {\n              image\n              caption\n              rubric\n              content\n            }\n          }\n          ... on PRISMIC_InsightBodyCallout {\n            type\n            primary {\n              callout\n              rubric\n              content\n            }\n          }\n        }\n        related_insights {\n          insight {\n            ... on PRISMIC_Insight {\n              _meta {\n                uid\n              }\n              title\n              feature_image\n              category {\n                ... on PRISMIC_Category {\n                  name\n                }\n              }    \n            }\n          }\n        }\n        get_in_touch_heading\n        social_title\n        social_description\n        social_image\n      }\n      allSite_settingss {\n        edges {\n          node {\n            get_in_touch_heading\n          }\n        }\n      }\n    }\n  }\n","id":"alEcZhIAAC0A7qLC","lang":"en-us","uid":"agentic-ai-what-it-means-for-communicators-and-bra","type":"insight","alternateLanguages":[],"cursor":"YXJyYXljb25uZWN0aW9uOjE4OQ==","paginationPreviousMeta":{"id":"ajwlaREAACcAVNfk","lang":"en-us","uid":"executive-media-training","type":"insight","alternateLanguages":[]},"paginationPreviousUid":"executive-media-training","paginationPreviousLang":"en-us","paginationNextMeta":{"id":"alZFBRIAACkA9n_D","lang":"en-us","uid":"what-nonprofits-can-learn-from-populist-messaging","type":"insight","alternateLanguages":[]},"paginationNextUid":"what-nonprofits-can-learn-from-populist-messaging","paginationNextLang":"en-us","lastQueryChunkEndCursor":"YXJyYXljb25uZWN0aW9uOjE3OQ=="}}}