Why isn't ebay doing anything to stop those scams?
Scam listings for Apple M3 Ultra devices priced around $1,000 are flooding eBay, posted by accounts with zero sales history, and the platform's AI moderation systems are failing to catch them. This matters because it reveals a fundamental gap between how eBay talks about marketpl...
A user identified as KillerMiller13, posting to Reddit's LocalLLaMA community, flagged what appears to be a coordinated wave of fraudulent eBay listings targeting buyers interested in high-end Apple silicon hardware. The post calls out eBay directly for leaving dozens of obvious scam listings active while real buyers bid on them, apparently unaware they are being defrauded. According to the LocalLLaMA community thread, the listings are easy to spot if you know what you are looking at, but eBay's automated systems are apparently not catching them.
Why This Matters
This is not a minor moderation hiccup. Dozens of scam listings from zero-history seller accounts, all offering an Apple M3 Ultra with 512GB storage for roughly $1,000, represent a systematic failure that eBay's AI enforcement tools should be built to catch. The M3 Ultra with 512GB of unified memory retails closer to $10,000 in legitimate markets, making the pricing anomaly enormous and obvious. eBay processed $73.2 billion in gross merchandise volume in 2024 according to its annual filings, and if its moderation infrastructure cannot flag a tenfold pricing discrepancy on a known high-value item posted by a brand-new account, the company has a serious trust problem that affects every buyer on the platform.
Daily briefing from 50+ sources. Free, 5-minute read.
The Full Story
The listings KillerMiller13 flagged follow a textbook fraud pattern. New seller accounts, zero completed transactions, and a premium technology item priced at a fraction of its real market value. An Apple M3 Ultra Mac Studio with 512GB of unified memory costs thousands of dollars through legitimate channels. For context, the Reddit post points out that even a 16TB SSD, a storage device alone, would cost more than these listings are asking for an entire high-performance computer. Buyers who complete purchases from these sellers will almost certainly receive nothing or receive a counterfeit item.
The problem is not that eBay lacks tools to address this. The company has been publicly investing in what it calls "Item Intelligence," a product initiative designed to handle regulatory compliance and content moderation through AI and machine learning. A 2025 job posting on eBay's careers page actively recruited a Lead Product Manager for Item Intelligence, with a mandate to "lead the product strategy to evolve our platform capabilities, translate regulatory and compliance problems into AI ML use cases and ensure eBay is a safe, trusted and easy to navigate marketplace." The job posting itself is an admission that current capabilities are not where they need to . eBay rolled out a new listing-report workflow last year, intended to make it easier for users to flag policy violations. Community feedback documented on the eBay seller forums was immediate and critical. Sellers and buyers reported that bad actors had already identified ways to evade the AI bots responsible for reviewing flagged content. The system that was announced as an improvement arrived already compromised in practical terms.
The failure extends beyond fraud listings. A documented incident from February 2026, covered by a seller-education YouTube channel, showed eBay's AI systems overturning a seller decision that was already correct. In that case, a seller shipped an item confirmed delivered on January 29th with full tracking documentation. The buyer opened an "Item Not Received" case. eBay reviewed the tracking evidence and closed the case in the seller's favor. Hours later, the buyer appealed, claiming no refund had been issued, despite never opening a return request and never providing return tracking. eBay's automated system then reversed its own prior correct decision and refunded the buyer anyway, forcing the seller into a manual appeal process. This is not a fringe case. It illustrates that eBay's AI is making inconsistent decisions that punish sellers who followed every rule correctly.
Adding to the irony, eBay made headlines in early 2026 by explicitly banning AI "buy for me" agents in a user agreement update. That announcement generated over 340 points of discussion on Hacker News, with the tech community debating whether the ban addressed any real problem. Critics pointed out the obvious contradiction: eBay is prohibiting external AI agents from interacting with its marketplace while its own internal AI agents are actively failing to protect buyers from human scammers running basic fraud schemes.
Key Details
- Dozens of eBay listings from zero-sales-history accounts offered Apple M3 Ultra 512GB devices for approximately $1,000 as of the Reddit post date.
- The legitimate market price for an M3 Ultra Mac Studio with 512GB of unified memory is approximately $10,000.
- eBay posted a job listing in 2025 for a Lead Product Manager for Item Intelligence to improve AI-driven compliance and moderation.
- eBay banned AI "buy for me" agents in early 2026, a move that received 340-plus points of discussion on Hacker News.
- A documented February 2026 seller case showed eBay's AI reversing a correctly decided dispute in favor of a buyer who provided no return tracking and filed no formal return request.
- eBay recorded $73.2 billion in gross merchandise volume in 2024.
What's Next
eBay's hiring push for Item Intelligence leadership in 2025 suggests the company plans to rebuild its moderation stack with a stronger AI and machine learning foundation, but hiring a product manager is a starting point, not a solution. Buyers should treat any eBay listing from a zero-feedback seller offering high-value technology at steep discounts as a near-certain fraud attempt until eBay demonstrates measurable improvement in its enforcement detection rates. Watch for eBay's next seller policy update, expected mid-2026, to signal whether the company intends to address new-account listing restrictions as a near-term mitigation.
How This Compares
Amazon faces similar fraud listing problems on its marketplace, particularly with third-party sellers on the Buy Box, but Amazon's scale of investment in automated enforcement and its verified seller programs provide at least some structural resistance to zero-history account fraud at this level. Facebook Marketplace, by contrast, has largely accepted that fraud is endemic to its peer-to-peer model and places responsibility on buyers. eBay occupies an uncomfortable middle ground: it markets itself as a trusted structured marketplace but operates with the open-access seller model of an informal platform.
The broader AI content moderation story is relevant here. Platforms including YouTube, Reddit, and X have all struggled with the same fundamental problem: AI systems trained to identify known bad patterns are easily defeated by bad actors who slightly modify their approach. What makes eBay's situation particularly notable is that the fraud signals in these M3 Ultra listings are not subtle. New account plus implausible pricing plus high-value item is a pattern that any basic risk-scoring model should flag. The failure suggests eBay's enforcement may not be optimized around user safety as a primary objective, which is a meaningful concern for a platform built on buyer trust.
Compare this to what is happening in AI-powered fraud detection tools more broadly. Fintech companies and payment processors have deployed multi-signal fraud detection that cross-references account age, pricing anomalies, category-specific baselines, and behavioral patterns in real time. eBay, with its resources and data history, should be capable of matching that standard. The fact that it demonstrably is not raises questions about internal priorities, not just technical capability. For more context on how AI moderation is evolving across platforms, check AI Agents Daily news.
FAQ
Q: How can I spot a scam listing on eBay? A: Look for three red flags together: a seller account with zero or near-zero completed transactions, a price that is dramatically lower than any other listing for the same item, and a high-value product like a premium laptop or phone. If all three are present, assume the listing is fraudulent and report it before moving . Q: Why doesn't eBay just remove listings from new sellers for expensive items? A: eBay's open marketplace model has historically prioritized seller accessibility, meaning new accounts can list immediately without verification delays. Changing that policy would reduce fraud but would also create friction for legitimate new sellers, which is a trade-off eBay has not been willing to make despite ongoing pressure from its seller and buyer communities.
Q: What should I do if I already paid for a suspicious eBay listing? A: Open an "Item Not Received" case through eBay's resolution center immediately and contact your payment provider or credit card company in parallel. eBay's buyer protection policy covers most purchases, but acting quickly improves your chances of a full refund. Document every step of the process including screenshots of the listing and your communications with the seller.
eBay's fraud problem is not new, but the M3 Ultra scam wave and the documented AI enforcement failures arriving in close succession make 2026 a critical year for the company to prove its moderation investments are producing real results rather than press releases. If Item Intelligence delivers meaningful improvements in detecting obvious fraud signals, it will be worth covering as a genuine turnaround story. Subscribe to the AI Agents Daily weekly newsletter for daily updates on AI agents, tools, and automation.
Get stories like this daily
Free briefing. Curated from 50+ sources. 5-minute read every morning.
