San José shelter warns of AI dog-photo euthanasia scam

San José Animal Care and Services is dealing with the fallout from an online scam that used AI-altered images of real shelter dogs to falsely suggest the animals were in their final hours before euthanasia. Coverage from KTVU on February 24, 2026, and ABC7 on February 25 said the posts, shared in a Facebook group called “Saving Shelter Dogs From Euthanasia,” prompted a flood of calls from distressed people around the country. Shelter leaders said the claims were false, the images had been manipulated, and at least some of the dogs highlighted in the posts had already been adopted. (ktvu.com)

The flashpoint appears to have been posts about two dogs, Pongo and Lumi. According to KTVU, staff first grew suspicious when one image showed human-like tears, which spokesperson Gilbert Martinez cited as evidence of digital alteration. Officials said Pongo had previously been in the shelter’s care, but was healthy, had no behavioral concerns, and was never at risk of euthanasia. In Lumi’s case, the city said the viral post was false and AI-generated; KTVU added that the circulating image was not even of the dog named in the post. (ktvu.com)

The misinformation landed in a particularly sensitive environment. ABC7 noted that San José’s shelter has already faced public criticism following a 2024 audit, making it more vulnerable to viral narratives that fit existing public anxieties about shelter euthanasia. Officials pushed back on those claims, saying the shelter does not euthanize for space and uses euthanasia only as a last resort for serious medical or behavioral issues. That distinction matters, because the scam appears to borrow real dog names and ID numbers to create posts that feel credible enough to spread quickly. (ktvu.com)

There are signs this isn’t an isolated incident. ABC7 reported that Ventura County Animal Services also saw one of its dogs featured in a doctored post with a similar plea, and that its phone room was flooded by confused callers before the dog was adopted. San José officials told reporters they had heard from other shelters experiencing the same pattern. That suggests a broader scam template: scrape legitimate shelter listings, heighten the emotional cues with AI editing, then drive engagement, donations, or both through third-party social media pages. The donation motive is not fully documented in every case, but ABC7 reported San José officials believed scammers were seeking donations, likes, or reposts. That last point is an inference about the likely business model behind the campaign, based on shelters’ descriptions of the posts and their effects. (abc7news.com)

Outside experts are framing the issue as emotional manipulation amplified by increasingly accessible AI tools. ABC7 quoted San Jose-based technology expert Ahmed Banafa calling it “emotional manipulation,” and saying AI is making misinformation more convincing. Animal welfare groups have been sounding related alarms. In March 2025, San Francisco SPCA warned that scammers impersonating shelters or hospitals may send altered images of pets, make threats, and demand money or sensitive information. The FBI has issued similar guidance, warning that lost-pet scammers may reuse or edit posted photos to make their claims seem legitimate. (abc7news.com)

Why it matters: For veterinary professionals, shelter veterinarians, and practice teams fielding worried calls from pet parents, this is more than a social media nuisance. False euthanasia claims can jam phone lines, pull staff away from patient care and intake work, and damage trust in legitimate medical and shelter decision-making. They can also intensify harassment risk for already stretched teams, especially when a shelter’s real policies are flattened into viral, emotionally charged content. The practical takeaway is that hospitals, shelters, and rescue partners may need clearer public-facing fraud protocols: verify through official websites, direct phone numbers, and in-person visits; avoid responding to third-party fundraising pleas without confirmation; and train frontline staff to recognize common scam cues, including edited images and urgency-driven messaging. (ktvu.com)

What to watch: The next step will likely be a mix of platform enforcement and local damage control. San José said it reported the Facebook page, but the broader question is whether social platforms move quickly enough when AI-manipulated animal content starts driving panic and possible fraud. In the near term, expect more shelters, clinics, and veterinary associations to publish scam advisories for pet parents, and to standardize messaging around how euthanasia decisions, adoption status updates, and emergency fundraising are communicated. If similar incidents keep spreading across shelters, this could become a wider reputational and operational issue for the companion animal sector, not just a local misinformation story. (ktvu.com)