AI "Obituary Pirates"

Security researchers have captured direct evidence of the harvesting speed. Sophos threat researchers documented one February 2024 case in which a single obituary was republished, AI-expanded, across six different fraudulent sites within 48 hours of the original posting. Scraping-to-republication can happen in under two days ^[3].

Sophos researchers found that generative AI was "likely used to expand a brief social media post into lengthy tribute text," producing pages with "obvious errors, inaccuracies, or fabrications" ^[3]. Separately, Futurism tested AI obituary-writing tools directly and found they "tend to fabricate details and use excessive flowery language" — when asked to write about a fictional person, the tool invented character traits never mentioned in the prompt ^[4]. The fabrication risk has been directly tested and demonstrated, not just alleged ^[3][4].

The fake sites rely on SEO poisoning — content and metadata engineered to outrank the real obituary in search results when people search a recently deceased person's name ^[3]. Google has acknowledged the problem by name: its March 2024 search spam-policy update explicitly targeted "expired websites repurposed as spam repositories by new owners and obituary spam." That update was part of a broader crackdown Google says cut low-quality, unoriginal content across search results by 40%, later revised up to 45% — a sitewide figure covering all forms of "scaled content abuse," not one isolated to obituary pages specifically ^[5]. Google has named obituary spam as an explicit target of its spam policy, but has not published a number for the effect on obituary sites alone ^[5].

Sophos traced the revenue model on the fake sites it examined: affiliate commissions from bundled antivirus-software subscriptions, redirects to e-dating and adult-content sites, and revenue from web-push-notification clickbait schemes — some pages also displayed fake virus-alert popups impersonating brands like McAfee and Windows Defender ^[3]. In the broader obituary-piracy business (not all of it AI-generated), platforms have built substantial revenue this way: Echovita alone generated roughly CA$5 million in 2020, including commissions on flowers, candles, and memorial-tree sales ^[1]. The money comes mainly from ad-tech, affiliate redirects, and scam-adjacent traffic — not primarily from donation solicitations ^[1][3].

Beyond financial exposure, the harm shows up in specific, documented cases. A sourced summary of obituary-piracy cases describes multiple unauthorized obituaries appearing within hours of a 2023 subway death, many containing false information — wrong age, wrong location, wrong cause of death ^[1]. In a separate case, a Los Angeles Times journalist discovered that pirated obituary sites had published an obituary for her — while she was alive ^[1]. The fabrications aren't limited to flattering invention; they include getting basic facts about a real death wrong ^[1].

Google's March 2024 spam-policy update is the platform response that can be verified in its own words: it named "obituary spam" explicitly as a target, folded into a broader policy against "scaled content abuse" — content mass-produced by automation, humans, or both ^[5]. Courts have also acted against obituary-piracy operators, though in cases that predate the current AI-generation wave: a 2018 lawsuit against a Canadian obituary-piracy site ended in a default judgment of roughly CA$20 million in damages after the site had posted over one million Canadian obituaries without permission ^[1]. Real legal consequences exist, but the enforcement actions specific to AI-generated obituary content that circulated in some 2024–25 coverage — an FTC consumer alert, state attorney general investigations, a dedicated FBI fraud category — could not be verified this pass and are not included here ^[1][5].

The core phenomenon is real and independently documented: AI tools are being used to scrape death notices and generate fabricated "tribute" content, monetized through ad-tech and affiliate schemes, and Google has named "obituary spam" as an explicit target of its search-spam enforcement ^{[1][2][3][4][5]}.

Several statistics that circulated in earlier coverage of this story could not be verified this pass, including a claimed National Funeral Directors Association survey figure, a specific family count from Consumer Reports, a detailed Washington Post family account, a February 2024 FTC consumer alert, state attorney general investigations in California, New York, and Texas, and a dedicated FBI Internet Crime Complaint Center category for "AI-enabled memorial fraud." None of the pages originally cited for these claims could be located this run; rather than replace them with guesses, they have been removed.

The scam is real and platforms are responding to it, but the most precise-sounding numbers attached to this story do not hold up under verification — treat any specific percentage or count in obituary-scam coverage with skepticism until it traces to a live, on-topic source.