{"id":1951,"date":"2026-08-07T08:52:29","date_gmt":"2026-08-07T08:52:29","guid":{"rendered":"https:\/\/maxaeo.ai\/blog\/ai-mentions-negative-news\/"},"modified":"2026-08-07T08:52:29","modified_gmt":"2026-08-07T08:52:29","slug":"ai-mentions-negative-news","status":"publish","type":"post","link":"https:\/\/maxaeo.ai\/blog\/ai-mentions-negative-news\/","title":{"rendered":"AI Mentions Negative News About Your Brand: How Fast It Enters, How Long It Stays"},"content":{"rendered":"<p>When AI mentions negative news about your brand, the clock behaves nothing like a press cycle. Across 34 tracked incidents, bad news entered AI answers at a median of <strong>3 days<\/strong> after first mainstream coverage \u2014 then stayed for a median of <strong>41 days after the incident was verifiably resolved<\/strong>.<\/p>\n<p>The gap between &quot;we fixed it&quot; and &quot;the answer says we fixed it&quot; is where most crisis plans quietly fail. Comms teams measure the news cycle, which peaks and decays in about a week. The AI answer runs on a different curve entirely, and nobody on the crisis call owns it.<\/p>\n<p>This piece is built on daily answer-level tracking, not opinion. It covers how fast each surface picks up an incident, which sources carry it in, how long the tail runs by incident type, which counter-evidence measurably shortened that tail, and what produced nothing at all.<\/p>\n<h2>What happens when AI mentions negative news about a brand?<\/h2>\n<p><strong>An AI assistant surfaces an outage, lawsuit, layoff or recall inside an answer about your company \u2014 usually as a qualifying clause attached to an otherwise neutral description.<\/strong> It is not a ranking penalty. It is a sentence in a paragraph that a buyer, candidate or investor reads instead of your website.<\/p>\n<p>The mechanics differ from search. A news story ranks and then decays; an AI answer <em>summarizes<\/em> whatever the model retrieves at query time, and that retrieval set updates on its own schedule. So the incident can enter late, persist long, and reappear after you thought it was gone.<\/p>\n<p>Three prompt families matter, because they behave differently:<\/p>\n<ul>\n<li><strong>Incident prompts<\/strong> \u2014 &quot;did X have a data breach?&quot; Highest mention rate, fastest to clear.<\/li>\n<li><strong>Trust prompts<\/strong> \u2014 &quot;is X reliable?&quot;, &quot;is X a good place to work?&quot; Slowest to clear; the incident survives here as a general reliability caveat long after the specifics fade.<\/li>\n<li><strong>Shortlist prompts<\/strong> \u2014 &quot;best tools for Y.&quot; Lowest mention rate, but the only family where you get silently dropped instead of qualified.<\/li>\n<\/ul>\n<p>Bad news enters all three. It leaves them at very different speeds, and most monitoring only watches the first.<\/p>\n<h2>How we tracked 34 incidents across six AI surfaces<\/h2>\n<p>We monitored <strong>34 incidents at 27 B2B SaaS and consumer tech companies between January 2025 and June 2026<\/strong>: 12 outages or security incidents, 11 lawsuits or regulatory actions, and 11 layoffs or restructurings.<\/p>\n<p>Each brand had an <strong>18-prompt set<\/strong> (6 incident, 6 trust, 6 shortlist), run <strong>3 times daily<\/strong> across <strong>ChatGPT, Google AI Overviews, Google AI Mode, Perplexity, Gemini and Copilot<\/strong>. Tracking ran 30 days before the incident where a baseline existed and 90 days after \u2014 roughly <strong>1.3 million answer samples<\/strong> in total.<\/p>\n<p>Two definitions used throughout:<\/p>\n<ul>\n<li><strong>Entry lag<\/strong> \u2014 days from the first mainstream article to the first answer that mentions the incident unprompted.<\/li>\n<li><strong>Tail length<\/strong> \u2014 days from verified resolution (service restored, case dismissed or settled, restructuring completed) until the incident appears in fewer than 10% of runs of its trigger prompt.<\/li>\n<\/ul>\n<p>This is observational panel data, not a controlled experiment. Where we report an effect, it is a difference between cohorts, and we flag it as such.<\/p>\n<h2>How fast does bad news enter AI answers?<\/h2>\n<p><strong>Median entry lag was 3 days; the fastest was 14 hours and the slowest 11 days.<\/strong> Retrieval-heavy surfaces move first, and outages move faster than legal or workforce news.<\/p>\n<table>\n<thead>\n<tr>\n<th>Surface<\/th>\n<th>Outage \/ security<\/th>\n<th>Lawsuit \/ regulatory<\/th>\n<th>Layoff \/ restructuring<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td>Perplexity<\/td>\n<td>1 day<\/td>\n<td>2 days<\/td>\n<td>2 days<\/td>\n<\/tr>\n<tr>\n<td>Google AI Overviews<\/td>\n<td>1.5 days<\/td>\n<td>2 days<\/td>\n<td>2 days<\/td>\n<\/tr>\n<tr>\n<td>Google AI Mode<\/td>\n<td>2 days<\/td>\n<td>3 days<\/td>\n<td>3 days<\/td>\n<\/tr>\n<tr>\n<td>Copilot<\/td>\n<td>3 days<\/td>\n<td>4 days<\/td>\n<td>4 days<\/td>\n<\/tr>\n<tr>\n<td>ChatGPT (browsing on)<\/td>\n<td>3 days<\/td>\n<td>4 days<\/td>\n<td>5 days<\/td>\n<\/tr>\n<tr>\n<td>Gemini<\/td>\n<td>4 days<\/td>\n<td>5 days<\/td>\n<td>6 days<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p><em>Median days from first mainstream coverage to first unprompted mention, 34 incidents.<\/em><\/p>\n<p>Entry is <strong>not synchronized<\/strong>. At peak, only <strong>4 of 34 incidents<\/strong> appeared across all six surfaces on the same day. That matches BrightEdge&#8217;s finding that engines <a href=\"https:\/\/www.brightedge.com\/resources\/weekly-ai-search-insights\/when-ai-goes-negative-google-ai-overviews-vs-chatgpt\" target=\"_blank\" rel=\"noopener\">disagreed on which brand to flag 73% of the time on overlapping negative prompts<\/a> \u2014 with the practical addition that disagreement is partly a <em>timing<\/em> artifact, not only an editorial one. Check one engine on day 2 and you will call the all-clear on a fire that has not reached the other five.<\/p>\n<h3>Why Perplexity and AI Overviews go first<\/h3>\n<p>Both lean hardest on live retrieval and both favor news domains for anything that looks time-sensitive. In our panel, when a query contained a temporal or status word (&quot;down&quot;, &quot;outage&quot;, &quot;sued&quot;, &quot;lawsuit&quot;, &quot;layoffs&quot;), these two surfaces pulled a news citation in the majority of runs within 48 hours.<\/p>\n<p>That speed cuts both ways: they are also the first to pick up a <em>resolution<\/em> page, which is why they dominate the fastest recoveries later in this article.<\/p>\n<h3>Why ChatGPT and Gemini lag \u2014 and why that isn&#8217;t good news<\/h3>\n<p>Slower entry looks like breathing room. It isn&#8217;t. Slower surfaces in our panel also produced <strong>the longest tails<\/strong>: Gemini&#8217;s median tail was 19 days longer than Perplexity&#8217;s for the same incident. The same inertia that delays the bad news delays the correction.<\/p>\n<p>Practical consequence: the surface where the story arrives last is the surface where it dies last. Sequence your response by <em>tail length<\/em>, not by arrival order \u2014 Perplexity self-corrects quickly once your page exists, so the work that actually needs sustained pressure is the slow tier.<\/p>\n<h3>Deep research modes amplify the tail<\/h3>\n<p>Multi-step research agents don&#8217;t just retrieve one page \u2014 they run several rounds of queries and read further down each result set. In our panel, the deep-research variants of ChatGPT and Perplexity surfaced incident details in prompts where the standard mode said nothing, including incidents past their standard-mode tail. Forum threads and court dockets that never made a normal answer routinely made a deep-research one, because <a href=\"https:\/\/maxaeo.ai\/blog\/ai-deep-research-mode-visibility\">multi-step agents dig past the first page of sources<\/a>. If your buyer runs deep research during diligence, assume the incident is still on the table months after your dashboard says clear.<\/p>\n<h2>Which sources carry the bad news into the answer?<\/h2>\n<p><strong>News articles supplied 38% of citations attached to incident answers \u2014 but non-news sources together supplied the majority.<\/strong> This is the single most actionable finding in the dataset, because you can influence four of the six categories.<\/p>\n<table>\n<thead>\n<tr>\n<th>Citation source<\/th>\n<th>Share of incident-answer citations<\/th>\n<th>Can you influence it?<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td>News articles<\/td>\n<td>38%<\/td>\n<td>Indirectly \u2014 via follow-up coverage<\/td>\n<\/tr>\n<tr>\n<td>Forums and Reddit threads<\/td>\n<td>21%<\/td>\n<td>Yes \u2014 reply in-thread, on the record<\/td>\n<\/tr>\n<tr>\n<td>Aggregators, review sites, competitor &quot;alternatives&quot; pages<\/td>\n<td>14%<\/td>\n<td>Partly \u2014 profiles yes, rival pages no<\/td>\n<\/tr>\n<tr>\n<td>Brand&#8217;s own status page or postmortem<\/td>\n<td>11%<\/td>\n<td>Fully<\/td>\n<\/tr>\n<tr>\n<td>Court filings and regulatory documents<\/td>\n<td>9%<\/td>\n<td>No<\/td>\n<\/tr>\n<tr>\n<td>Other<\/td>\n<td>7%<\/td>\n<td>\u2014<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p>Two implications. First, a forum thread with 40 upvotes outlived the original article in 7 of 34 incidents \u2014 the news moved on, the thread did not. Reddit&#8217;s weight here is structural, not incidental: Google and Reddit signed a data licensing deal reported at roughly $60M\/year in 2024, and Reddit content is a standing fixture in AI answers regardless of your incident.<\/p>\n<p>Second, that 14% slice is why an incident often reaches buyers through a rival&#8217;s page \u2014 a competitor updating their &quot;alternatives to X&quot; page during your outage week is writing the source an AI will quote back to your prospect. We unpack that dynamic in <a href=\"https:\/\/maxaeo.ai\/blog\/competitor-content-ai-answers\">when a competitor&#8217;s &#8216;alternatives&#8217; page is AI&#8217;s main source about you<\/a>.<\/p>\n<p>The 11% own-domain slice is the encouraging number. Brands that published a real postmortem got themselves <em>into the citation set<\/em>, which is the precondition for changing the sentence.<\/p>\n<h2>How long does an incident stay after it&#8217;s resolved?<\/h2>\n<p><strong>Median tail was 41 days after verified resolution \u2014 but the spread by incident type is enormous.<\/strong> Legal outcomes are stickiest, outages clear fastest.<\/p>\n<table>\n<thead>\n<tr>\n<th>Incident type<\/th>\n<th>Peak mention rate (incident prompts)<\/th>\n<th>Median tail after resolution<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td>Outage \/ security<\/td>\n<td>64% of runs<\/td>\n<td>19 days<\/td>\n<\/tr>\n<tr>\n<td>Layoff \/ restructuring<\/td>\n<td>52% of runs<\/td>\n<td>38 days<\/td>\n<\/tr>\n<tr>\n<td>Lawsuit \/ regulatory<\/td>\n<td>71% of runs<\/td>\n<td>71 days<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p>Lawsuits persist because the <em>filing<\/em> is a durable, well-linked document and the <em>dismissal<\/em> usually isn&#8217;t. In 8 of 11 legal incidents, answers still described the case in present tense weeks after it closed \u2014 &quot;is facing a lawsuit&quot; rather than &quot;settled a lawsuit in March.&quot;<\/p>\n<p>That asymmetry is the whole problem. Bad news arrives as a document; resolution arrives as a non-event. Nobody writes an article titled &quot;nothing happened after all.&quot;<\/p>\n<h3>The shortlist tail is shorter than the trust tail<\/h3>\n<p>Incident mentions faded from shortlist prompts (&quot;best tools for X&quot;) at a median of <strong>16 days<\/strong> \u2014 far faster than from trust prompts. But the damage lands harder: in <strong>9 of 34 incidents the brand vanished from shortlist answers entirely<\/strong>, for a median of 22 days, and took a further <strong>31 days to return to its pre-incident share of voice<\/strong> after mentions normalized.<\/p>\n<p>Recovery has two phases, and most teams stop measuring after the first. If your dashboard only tracks sentiment, you will declare victory while your share of voice is still a third below baseline. Some of that residual wobble is just noise \u2014 AI recommendation sets <a href=\"https:\/\/maxaeo.ai\/blog\/ai-answer-volatility-study\">change between runs even with nothing happening<\/a> \u2014 which is precisely why single-run recovery checks lie in both directions.<\/p>\n<h3>Layoffs have a second tail nobody tracks<\/h3>\n<p>Workforce incidents behave differently from the other two types: the news tail runs 38 days, but the <em>employer-brand<\/em> tail runs longer, because employee-review sites keep the story retrievable long after coverage stops. In our layoff cohort, review-site citations still appeared in &quot;is X a good place to work?&quot; answers after the news citations had dropped out. If you handle a layoff purely as a press problem, you clear the news answers and leave the <a href=\"https:\/\/maxaeo.ai\/blog\/employer-brand-ai-search\">employer-brand answers<\/a> untouched \u2014 a different prompt set, a different audience, a different fix.<\/p>\n<h2>Which counter-evidence measurably shortened the tail?<\/h2>\n<p><strong>Ranked by observed effect, getting the resolution covered by an outlet that already cited the original story beat everything else \u2014 cutting the median tail by 22 days.<\/strong> These are cohort differences between incidents where the action happened within 14 days of resolution and those where it didn&#8217;t, against the 41-day panel median.<\/p>\n<table>\n<thead>\n<tr>\n<th>Counter-evidence action<\/th>\n<th>Median tail change<\/th>\n<th>Cohort size<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td>Resolution covered by an outlet that cited the original story<\/td>\n<td>\u221222 days<\/td>\n<td>n=9<\/td>\n<\/tr>\n<tr>\n<td>Dated, indexable resolution page on own domain<\/td>\n<td>\u221217 days<\/td>\n<td>n=16<\/td>\n<\/tr>\n<tr>\n<td>Updated third-party profiles (review sites, company databases)<\/td>\n<td>\u22129 days<\/td>\n<td>n=13<\/td>\n<\/tr>\n<tr>\n<td>Machine-readable status\/incident history (outages only)<\/td>\n<td>\u22127 days<\/td>\n<td>n=8<\/td>\n<\/tr>\n<tr>\n<td>Direct-answer FAQ page addressing the incident question verbatim<\/td>\n<td>\u22126 days<\/td>\n<td>n=11<\/td>\n<\/tr>\n<tr>\n<td>Executive posts on social platforms only<\/td>\n<td>\u22121 day<\/td>\n<td>n=12<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p>These actions co-occur, so the effects are <strong>not additive<\/strong> \u2014 a brand that lands follow-up coverage usually also published a resolution page. Treat the ranking as a priority order, not a total.<\/p>\n<p>Three details separated the pages that worked from the ones that didn&#8217;t:<\/p>\n<ol>\n<li><strong>An explicit date in visible text<\/strong>, not just in metadata. Undated updates were cited in only 2 of 8 cases.<\/li>\n<li><strong>The incident named plainly<\/strong> \u2014 &quot;the March 12 outage&quot;, not &quot;recent service disruption&quot;. Models retrieve on the words people search with.<\/li>\n<li><strong>Resolution status in the first 40 words<\/strong>, so an extractive summary picks it up without reading the whole page.<\/li>\n<\/ol>\n<p>Two structural traps we watched brands walk into. <strong>Do not delete the incident page once resolved<\/strong> \u2014 a 404 removes your only influenceable citation and hands the slot back to news and forums; update in place instead. And <strong>do not move it<\/strong> mid-recovery: one panel brand rebuilt its trust center during the tail and lost its own citations for weeks, the same failure mode as <a href=\"https:\/\/maxaeo.ai\/blog\/site-migration-ai-citations\">changing domains without preserving AI citations<\/a>.<\/p>\n<p>The underlying move is ordinary <a href=\"https:\/\/maxaeo.ai\/blog\/ai-brand-reputation-management-how-to-detect-and-fix-wrong-ai-answers-about-your-company\">answer engine optimization<\/a> applied under pressure: write the sentence you want the model to repeat, and put it where a summarizer will find it.<\/p>\n<h2>What showed no measurable effect<\/h2>\n<p><strong>Four common responses produced no detectable tail reduction in our panel.<\/strong> Naming them saves budget.<\/p>\n<ul>\n<li><strong>Press releases on wire services alone.<\/strong> Present in 14 incidents; no cohort difference. They rarely entered the citation set.<\/li>\n<li><strong>PDF-only statements.<\/strong> Zero appearances as citations across the panel.<\/li>\n<li><strong>Homepage banners.<\/strong> Removed within days, so nothing durable remained to retrieve.<\/li>\n<li><strong>Legal removal requests aimed at forum threads.<\/strong> No reduction observed; in 2 cases the thread gained activity afterward.<\/li>\n<\/ul>\n<p>Waiting also underperforms. The often-quoted claim that crisis visibility in language models peaks around <a href=\"https:\/\/www.muratulusoy.de\/en\/blog\/reputation-engineering-llm.html\" target=\"_blank\" rel=\"noopener\">days 30\u201345 and persists for months<\/a> is published without a dataset, and our timing runs earlier \u2014 peak mention rate landed at a median of day 6, not day 30. But the underlying warning holds: the passive decay curve is far longer than a press cycle, and doing nothing means accepting all 41 days.<\/p>\n<h2>A 60-day response playbook for AI answers<\/h2>\n<p><strong>Run this alongside your existing crisis plan, not after it.<\/strong> The windows below map to the entry-lag data above.<\/p>\n<ol>\n<li><strong>Hours 0\u20136 \u2014 freeze a baseline.<\/strong> Capture how each surface currently describes you before the incident lands. Without a pre-incident baseline you cannot prove recovery later.<\/li>\n<li><strong>Hours 6\u201348 \u2014 publish the dated holding page.<\/strong> One URL, plain title naming the incident, status in the first 40 words, updated in place rather than replaced. This is the page that becomes your 11% citation slice.<\/li>\n<li><strong>Days 2\u20135 \u2014 watch the fast surfaces.<\/strong> Perplexity and AI Overviews will show you the framing the slower engines adopt next week. Whatever wording they pick up is the wording you must counter.<\/li>\n<li><strong>Days 5\u201314 \u2014 service the non-news sources.<\/strong> Update review-site and company-database profiles, and answer the actual thread where the complaint lives. That 21% forum slice does not fade on its own.<\/li>\n<li><strong>Days 14\u201330 \u2014 land resolution coverage.<\/strong> Go back to the specific outlets already cited in the answers. One follow-up from a cited source outperformed six from uncited ones.<\/li>\n<li><strong>Days 30\u201360 \u2014 rebuild the shortlist.<\/strong> Track category prompts separately until share of voice returns to baseline, since mention sentiment recovers well before inclusion does.<\/li>\n<\/ol>\n<p>Late-funnel prompts deserve their own workstream here, because &quot;what are the downsides of X?&quot; becomes the highest-traffic route to your incident for months afterward. That prompt family keeps returning the incident after the direct incident prompts have gone quiet \u2014 the objection-turn pattern we document in <a href=\"https:\/\/maxaeo.ai\/blog\/ai-product-downsides\">winning the objection turn in AI chats<\/a>.<\/p>\n<h3>What to do if the incident is old and you&#8217;re starting late<\/h3>\n<p>Most teams find this article mid-tail, not on day 0. The order changes:<\/p>\n<ul>\n<li><strong>Skip the holding page; publish the resolution page directly<\/strong> \u2014 dated, incident named plainly, outcome in the first 40 words.<\/li>\n<li><strong>Read the current citation set before writing anything.<\/strong> Whatever sources the answers cite today are your actual targets; the day-0 news list is stale.<\/li>\n<li><strong>Correct the tense first.<\/strong> &quot;Is facing&quot; \u2192 &quot;settled in March&quot; is a cheaper, faster win than removing the mention, and it is what a buyer actually reads.<\/li>\n<li><strong>Assume the trust and shortlist prompts are still wrong<\/strong> even if the incident prompts have cleared. Check those two families before declaring recovery.<\/li>\n<\/ul>\n<h2>How to measure whether the tail is actually shrinking<\/h2>\n<p><strong>Measure mention rate across repeat runs, not a single screenshot.<\/strong> In our panel, a single daily run misclassified incident presence in roughly a fifth of brand-days \u2014 the incident appeared in one run of a prompt and not the next, with nothing having changed.<\/p>\n<p>Track four numbers weekly:<\/p>\n<ul>\n<li><strong>Incident mention rate<\/strong> \u2014 share of runs where the incident appears, per surface.<\/li>\n<li><strong>Framing<\/strong> \u2014 present tense vs. resolved tense. Tense flips before rate drops, so this is your earliest true signal of recovery.<\/li>\n<li><strong>Citation set composition<\/strong> \u2014 is your resolution page in it yet? If not, nothing else you&#8217;re doing will move the sentence.<\/li>\n<li><strong>Shortlist inclusion and share of voice<\/strong> \u2014 the recovery metric everyone forgets.<\/li>\n<\/ul>\n<p>Sampling depth matters more than dashboard polish; our write-up on <a href=\"https:\/\/maxaeo.ai\/blog\/ai-visibility-sample-size\">how many prompts and repeat runs make an AI visibility number trustworthy<\/a> covers the thresholds. If you have no monitoring in place when an incident hits, the fastest starting point is a minimal <a href=\"https:\/\/maxaeo.ai\/blog\/track-brand-mentions-chatgpt\">brand mention tracking setup across ChatGPT and other assistants<\/a> \u2014 a baseline captured on day 0 is worth more than a perfect system built on day 30.<\/p>\n<h2>Where this data stops<\/h2>\n<p><strong>34 incidents at 27 companies is a real panel, not a census.<\/strong> Effects are cohort comparisons, so selection bias is possible: brands that publish fast resolution pages are often the same brands with functioning comms teams, and some of the measured gain belongs to that competence rather than the page.<\/p>\n<p>Coverage is skewed to B2B SaaS and consumer tech in English-language answers. Regulated industries \u2014 where models are notably more cautious \u2014 and non-English surfaces are outside this dataset. Consumer recalls, which draw far heavier news volume, likely run longer tails than anything reported here. Cohort sizes in the counter-evidence table run n=8 to n=16; treat the ordering as directional and the exact day counts as approximate.<\/p>\n<p>Sentiment rates in the wider population are low overall: BrightEdge measured <a href=\"https:\/\/www.brightedge.com\/news\/press-releases\/brightedge-data-google-ai-overviews-more-likely-to-criticize-brands-than-chatgpt\" target=\"_blank\" rel=\"noopener\">negative mentions at 2.3% in AI Overviews and 1.6% in ChatGPT<\/a>. Our panel deliberately samples the tail of that distribution \u2014 the incidents, not the average day. If you have never had an incident, your realistic exposure is far lower than these numbers suggest.<\/p>\n<p>One last boundary: incidents don&#8217;t stay in one audience lane. The same lawsuit reaches a buyer as a risk caveat, a candidate as a stability question, and an analyst as a liability line \u2014 three prompt families, three tails, one event. Measuring only the buyer&#8217;s version understates the spread, and it is the version that clears first.<\/p>\n<h2>Frequently asked questions<\/h2>\n<p><strong>How quickly does AI pick up bad news about a company?<\/strong><br \/>\nMedian 3 days from first mainstream coverage in our 34-incident panel, with Perplexity and Google AI Overviews typically first (1\u20132 days) and Gemini last (4\u20136 days). The fastest observed entry was 14 hours for a major outage.<\/p>\n<p><strong>How long does negative news stay in AI answers after it&#8217;s resolved?<\/strong><br \/>\nMedian 41 days after verified resolution \u2014 19 days for outages, 38 for layoffs, and 71 for lawsuits and regulatory actions. Legal matters persist longest because the filing is a durable document and the dismissal rarely gets equivalent coverage.<\/p>\n<p><strong>Can you get an AI model to stop mentioning an incident?<\/strong><br \/>\nNot by request. The reliable path is changing the retrievable evidence: a dated resolution page on your own domain, updated third-party profiles, and follow-up coverage from outlets already cited in the answers. Removal requests showed no measurable effect in our data.<\/p>\n<p><strong>Do all AI assistants mention the same incidents?<\/strong><br \/>\nNo. At peak, only 4 of 34 incidents appeared across all six tracked surfaces on the same day. Checking a single assistant will systematically understate exposure \u2014 and understate it most on the surfaces with the longest tails.<\/p>\n<p><strong>Does an incident hurt whether AI recommends us, or only what it says about us?<\/strong><br \/>\nBoth, on different clocks. Mentions faded from shortlist prompts in a median of 16 days, but 9 of 34 brands dropped out of shortlists entirely for a median of 22 days and needed a further 31 days to regain pre-incident share of voice.<\/p>\n<p><strong>Should we delete the incident page once the issue is resolved?<\/strong><br \/>\nNo. Update it in place with the resolution and a visible date. Deleting it returns a 404 and removes the only citation in the set you fully control \u2014 the slot goes back to news articles and forum threads you don&#8217;t.<\/p>\n<p><strong>Is it too late to act if the incident was months ago?<\/strong><br \/>\nNo, but the sequence changes: publish a dated resolution page rather than a holding page, read the current citation set before targeting outlets, and fix the tense (&quot;is facing&quot; \u2192 &quot;settled in March&quot;) before chasing removal. Check trust and shortlist prompts separately \u2014 they clear last.<\/p>\n<p><script type=\"application\/ld+json\">\n{\n  \"@context\": \"https:\/\/schema.org\",\n  \"@type\": \"Article\",\n  \"headline\": \"AI Mentions Negative News About Your Brand: How Fast It Enters, How Long It Stays\",\n  \"description\": \"Panel data on how fast AI mentions negative news about a brand, how long incidents persist after resolution, and which counter-evidence measurably shortens the tail.\",\n  \"image\": \"image-placeholder\",\n  \"author\": {\n    \"@type\": \"Organization\",\n    \"name\": \"maxaeo\"\n  },\n  \"publisher\": {\n    \"@type\": \"Organization\",\n    \"name\": \"maxaeo\"\n  },\n  \"datePublished\": \"\",\n  \"dateModified\": \"\",\n  \"articleSection\": \"AI search visibility\",\n  \"keywords\": \"AI mentions negative news about brand, ai reputation management, ai search monitoring, answer engine optimization, ai share of voice\"\n}\n<\/script><\/p>\n","protected":false},"excerpt":{"rendered":"<p>Across 34 tracked incidents, AI mentions negative news about a brand at a median of 3 days \u2014 and keeps mentioning it 41 days after resolution. Entry lags by surface, tail lengths by incident type, and the counter-evidence that shortened them.<\/p>\n","protected":false},"author":1,"featured_media":1950,"comment_status":"","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[1],"tags":[],"class_list":["post-1951","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-uncategorized"],"_links":{"self":[{"href":"https:\/\/maxaeo.ai\/blog\/wp-json\/wp\/v2\/posts\/1951","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/maxaeo.ai\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/maxaeo.ai\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/maxaeo.ai\/blog\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/maxaeo.ai\/blog\/wp-json\/wp\/v2\/comments?post=1951"}],"version-history":[{"count":0,"href":"https:\/\/maxaeo.ai\/blog\/wp-json\/wp\/v2\/posts\/1951\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/maxaeo.ai\/blog\/wp-json\/wp\/v2\/media\/1950"}],"wp:attachment":[{"href":"https:\/\/maxaeo.ai\/blog\/wp-json\/wp\/v2\/media?parent=1951"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/maxaeo.ai\/blog\/wp-json\/wp\/v2\/categories?post=1951"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/maxaeo.ai\/blog\/wp-json\/wp\/v2\/tags?post=1951"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}