AI Is Flooding Newsrooms With Clichés. Here’s How And Why You Should Clean It Up

<div><p><strong>Try this</strong>: Search your inbox for the word “landscape.” Unless you enjoy discussing your gardening with coworkers, you’ll likely find an assortment of AI-generated newsletters and emails stuffed with lazy language phrases like “AI landscape” or “ever-evolving landscape.” For reference, here’s my inbox:</p><div><img width="1024" src="https://tvnewscheck.com/wp-content/uploads/2025/06/landscape-inbox-Acca…; alt="">Image via Jon Accarrino</div><p>But these clichés aren’t just in your inbox –  they’re probably also in your newsroom’s content. AI-generated cliches can appear everywhere from broadcast script summaries and social posts to expert quotes in news stories. As large language model-powered productivity tools become more common in newsrooms,  AI’s unfortunate <a href="https://slashdot.org/story/22/10/16/2024253/why-mastering-language-is-s… for clichés</a> is also coming along for the ride. Here’s why that matters and what news orgs can do about it.</p><p>For news organizations, clichés can be credibility killers. When readers encounter phrases like “buckle up” or “breakneck speed” in news content, it can trigger skepticism because these expressions sound out of place in serious journalism.</p><p>Research from Stanford University <a href="https://stanforddaily.com/2025/03/13/new-software-allows-ai-detection/"…; that readers could identify AI-generated content with high accuracy, often citing “formulaic language” and “overused phrases” like cliches as reasons to question authenticity. When audiences suspect content is AI-written, trust in a news brand can suffer.</p><p>Even if a reporter avoids using AI tools themselves, news subjects and experts may be using AI to generate their quotes. When experts use tools like ChatGPT to create statements, it makes it harder for both newsrooms and audiences to distinguish real expertise from AI-generated commentary. Unfortunately, AI-generated quotes can also contribute to misinformation and reputation sabotage. Recent <a href="https://www.cbsnews.com/news/reporter-artificial-intelligence-fake-quot…; have shown how AI can be used to invent plausible but <a href="https://learningenglish.voanews.com/a/us-reporter-used-ai-to-create-fal… quotes</a> by <a href="https://www.provokemedia.com/latest/article/perspective-fake-experts-ar… experts</a>. Worse yet, sometimes fake AI-generated quotes are attributed to real people. Or even <a href="https://www.msn.com/en-us/movies/news/clint-eastwood-said-he-never-gave… interviews</a>.</p><p>So, why does AI love cliches? AI doesn’t write — it predicts. Large language models like ChatGPT are trained on massive amounts of data. As a result, they default to statistically probable phrases like “game changer” or “in the wake of.”</p><p>“Artificial Intelligence can be amazing for improving productivity,” says <a href="https://www.linkedin.com/in/mrevanyoung/">Evan Young</a>, COO of <a href="https://heynota.com">NOTA</a&gt;. “But if you want to remain trusted by and connected to your audience, it’s always vital to ensure your voice, tone, and style come across as authentic. Unfortunately, most out of the box AI tools struggle with this.”</p><p>This statistical method leads to predictable patterns in AI output:</p><ul><li><strong>A feedback loop of mediocrity</strong>: AI favors an overuse of intensifiers, such as “game-changing” and “revolutionary,”  because they’re common in its training data.</li> <li><strong>Context Blindness</strong>: AI lacks the nuanced judgment to know when a phrase like “here’s the rub” is inappropriate or out of place in a news report.</li> <li><a href="https://x.com/MelMitchell1/status/1661462021571764228"><strong>Regressi… to the meh</strong></a>: With AI content now comprising <a href="https://originality.ai/ai-content-in-google-search-results">19% of Google search results</a> (as of Jan 2025), future models risk amplifying the problem. If new AI models train on AI-generated content full of lazy language and cliches, then it will create a self-perpetuating cycle of linguistic mediocrity.</li> <li><strong>Loss of distinct voice</strong>: Overreliance on AI can erode a publication’s unique style, making content blend into the digital noise.</li></ul><p>“You can’t cut through the noise by sounding like everyone else,” says <a href="https://www.linkedin.com/in/lee-zurik-a736525/">Lee Zurik</a>, SVP of News Strategy and Innovation at <a href="https://www.gray.tv/">Gray Media</a>. “Newsrooms need to edit AI-generated content with the same care we give any script.” </p><p>While AI tools like NOTA already limit lazy language from being used in newsroom outputs, it’s helpful to proactively teach staff how to identify and deal with cliches and trite expressions. Here are some tips:</p><ul><li>Create custom cliche detection apps in your AI enterprise platform, for example this <a href="https://studio.ordodigital.com/Cliche_Killer_F32O3">Cliché Killer app</a>, or use tools like Originality.ai, GPTZero, or Copyleaks to detect AI-generated content and flag overused terms</li> <li>Train teams on how to <a href="https://methodshop.com/manual-ai-generated-text-detection/">manually identify AI-generated content</a></li> <li>Run cliché density reports to track and improve your newsroom’s content quality</li></ul><ul><li>Use expert database tools, like <a href="http://rolli.ai/">Rolli</a&gt;, that verify identities of experts</li> <li>Use AI detection tools to scan expert quotes for lazy language patterns and cliches or interview sources via phone</li> <li>Maintain a newsroom-specific list of forbidden phrases, updating it regularly as new AI-generated clichés are identified</li> <li>Resources like the “<a href="https://docs.google.com/spreadsheets/d/1Qb7FJji4Rtx6thCfFQ3Wl0vnZ0tdbTQ… Blacklist for AI Chatbots</a>” provide a crowdsourced starting point. This is a free resource listing the most common clichés overused by chatbots created by <a href="https://ordodigital.com/">Ordo Digital</a>, <a href="https://mediacopilot.ai/">The Media Copilot</a> and <a href="https://asknews.app/">AskNews</a>.</li></ul><p>For newsrooms, it’s important to keep a human-in-the-loop with all AI processes that create content. </p><ul><li><strong>The “so what” test</strong>: Replace vague buzzwords with concrete facts or measurable impact. For example, instead of “Our new weather coverage is a <em>game changer</em>,” say “WFIR increased audience retention by 17% using a new weather coverage format in Q1 tests.”</li> <li><strong>Prompt training</strong>: Use precise prompts when generating AI text. Instead of “Write a press release,” try “Draft a 300-word announcement for [Product] including two customer outcomes and one hard metric, avoiding buzzwords.”</li> <li><strong>Style-guide lockouts:</strong> Add a “Forbidden Phrases” appendix to editorial guidelines to discourage cliché usage.</li> <li><strong>Get direct quotes from sources</strong>: “Direct conversations, by phone or in person, help ensure you’re quoting real, vetted experts, not AI-generated personas,” says <a href="https://www.linkedin.com/in/nicktoso/">Nick Toso</a>, CEO of <a href="http://rolli.ai">Rolli</a&gt;. “As fake experts and AI-generated quotes become more common, newsrooms should prioritize source verification and real-time interviews to protect the authenticity of their reporting.”</li> <li><strong>Positive reinforcement</strong>: Reward team members who identify and eliminate clichés to foster a culture of originality.</li></ul><p><strong>Raise awareness around clichés</strong>: Designer Thomas Smith created laser-etched “<a href="https://medium.com/the-generator/i-turned-chatgpts-horrible-writing-cli… Coasters</a>” featuring AI’s most overused phrases as a daily reminder to prioritize originality. Tactics like this are a great way to raise awareness around lazy language with your team.</p><div><img width="720" src="https://tvnewscheck.com/wp-content/uploads/2025/06/cliche-coasters.jpg&…; alt="">‘AI Cliche Coasters’ via <a href="https://medium.com/the-generator/i-turned-chatgpts-horrible-writing-cli… Smith</a></div><p>By eliminating clichés, media organizations are not just improving their product; they are actively preserving authenticity and original human writing, which are increasingly becoming premium assets on an AI-powered internet. </p><p>“Using AI isn’t the differentiator anymore, using it wisely is,” says <a href="https://www.linkedin.com/in/strategicandgoaldriven/">Mary Rogers</a>, SVP of Sales and Marketing at <a href="https://futurimedia.com/">Futuri</a&gt;. “The newsrooms that win won’t be the fastest, they’ll be the ones who combine predictive intelligence with human judgment to craft meaningful stories, faster, for every platform their audience lives on.”</p><p>For media companies, minimizing AI clichés leads to some real advantages:</p><ol><li><strong>Improved trust</strong>: Awkward phrases and overused clichés can cause readers to question authenticity and erode confidence in news outlets that overlook them.  </li> <li><strong>Unique style</strong>: Reducing lazy language helps preserve each publication’s distinct voice as AI usage continues to increase with news orgs.  </li> <li><strong>Better engagement</strong>: Clear, jargon-free writing is more enjoyable and less boring or generic.</li></ol><p><em>The Atlantic</em>‘s editor <a href="https://www.linkedin.com/in/jeffrey-goldberg-648946347/">Jeffrey Goldberg</a> addressed the issue with cliches when <a href="https://www.theatlantic.com/technology/archive/2023/02/ai-chatbots-clic…; his team about AI usage: Nobody comes to <em>The Atlantic</em> to be ‘bored and annoyed’. Goldberg understood that clichés don’t just irritate readers – they can be a signal of lazy journalism that fails to deliver the original insight and careful analysis readers expect.</p><p>AI is here to stay, but its flaws don’t have to be. By auditing AI outputs, training teams to mitigate lazy language and prioritizing precision over predictability, newsrooms can take advantage of AI’s efficiency without sacrificing quality. The goal isn’t to reject AI, but to refine it because in an industry built on trust, the words we choose still matter most.</p><p><em>Disclosure: Gray Media is a client of Ordo Digital.</em></p></div>