跳到正文
Analytics India Magazine· Kileuna Haralu·· 1 小時前AI 評分54

從 Raye 到 Coldplay:AI 黃牛如何推高票價、令歌迷難以購票

From Raye to Coldplay, How AI Scalpers are Pricing You Out of Concert Tickets

AI 導讀

文章梳理 AI 與傳統自動化工具如何助長演唱會黃牛及轉售詐騙,並指出平台須分辨真實持票者與獲授權的自動化交易。Raye 在 Royal Albert Hall 的演唱會登記收到 1.8 million 次申請,但只有 5,000 個座位;Coldplay 去年在印度開唱時,近 13 million 名歌迷爭購 150,000 張門票。

正文

British artiste Raye may be poised to dominate the Grammys this year, but the Where is My Husband! singer is overwhelmed by a set of fans she wouldn’t have ever imagined to encounter: AI bots. The registration for her concert at London’s Royal Albert Hall attracted 1.8 million sign-ups for just 5,000 seats. 

"Sadly, our sign-up process was overwhelmed by scalpers, using AI bots to create hundreds of thousands of fake registrations," Raye wrote on X. "It's a new world we are living in now; I have never witnessed bots on this scale." Her team subsequently decided to manually review registrations.

The concerts and ticketing industry, like the rest of the world, is beginning to reckon with the AI era. 

Earlier, scalper bots could be blocked by simple IP blocks and “traffic light” tests. Now, scalpers use thousands of generated digital identities, continuously scraping through resale demand and seat availability to adjust markups.

For India’s ₹20,000 crore live and concert market, speed-ticket scalping is sparking fears over new kinds of fraud and revenue loss, and has also prompted many platforms to up their anti-bot defences.

AI-Driven Scalping Runs on Existing Tactics

When Coldplay came to India last year, BookMyShow saw an unprecedented demand for tickets, with nearly 13 million fans vying for 150,000 tickets available. A large number of tickets appeared on popular resale platforms such as Viagogo and Gigsberg, commanding astronomical markups. The culprit was bot-driven ticket scalping, where malicious operators used scalper bots trained to mimic human behaviour to bypass checks.

“Most ticket scalping in India still runs on old methods: scripts that check out faster than a human can, lots of accounts on different phone numbers, and people hired to wait in online queues. None of that needs AI,” Aniket Chaudhary, Founder of secondary ticket marketplace Reticket, tells AIM.

Conventional bots can refresh availability, navigate online queues, select seats, and submit checkout requests using predefined instructions. 

“The real issue is that ticketing systems were originally designed around human interaction, while automated systems can operate at a completely different speed and scale,” says Anuj Gupta, Founder and CEO of AI research lab KiteFishAI.

But AI-generated codedoes not automatically make a bot an AI agent. 

“A bot written with the help of ChatGPT is not necessarily an AI agent,” Gupta cautions. Much of the activity involved in ticket buying can still be handled by conventional scripts, with AI acting as an additional layer for developing or adapting them.

The technology also changes how fraudulent listings are created. Fake booking confirmations can reproduce familiar logos, booking IDs, and QR codes, supported by convincing descriptions and conversations with supposed buyers.

“Anyone can now make a booking confirmation that looks exactly like a real one in about a minute,” Chaudhary notes.

This makes visual checks less dependable. Events discovery platform MeraEvents identifies bulk registrations, duplicate accounts, unauthorised resale, and copied booking confirmations as potential risks. However, it cautions that “automation alone does not establish AI involvement”. 

The Bigger Risk

The problem now is whether ownership and usage of sold tickets can be tracked. A screenshot or PDF cannot establish that a seller still owns a ticket or has the right to transfer it.

“The bigger problem, though, is real tickets. Generally, more scams involve someone holding a genuine ticket than an AI-made fake,” Chaudhary explains.

Reticket uses zero-knowledge Transport Layer Security (zkTLS), through Reclaim Protocol, to obtain cryptographic proof that booking details came directly from the issuing platform's servers and were not altered. The verification checks the event, payment, and booking status. Each ticket can also be listed only once, and a listing is locked when a buyer begins payment.

Reticket holds sellers' payouts until 72 hours after the event. Where a transfer or wristband delivery is delayed, sellers must set up a refundable UPI mandate deposit ranging from ₹499 to ₹4,999 to compensate the buyer.

Gupta argues that ticketing platforms need to treat tickets as verifiable digital assets rather than images or documents. “Every ticket should be associated with a valid transaction and a unique cryptographic identifier that can be verified against the issuing platform,” he says. Dynamically refreshing QR codes, digitally signed ticket credentials, and controlled transfers could also make screenshots and duplicate tickets harder to reuse.

The responsibility does not rest with the technology platform alone. Event organisers need to communicate official booking channels, establish transfer and entry rules, and coordinate with ticketing and on-ground teams when discrepancies arise, MeraEvents notes.

The Resale Journey

A resale marketplace may verify a ticket before listing it, but it cannot always monitor what happens to that ticket on the original platform afterwards. This creates a gap between proving ownership at the time of resale and confirming that the ticket remains valid until entry.

“Our verification already works without any formal tie-up, because sellers prove ownership from their own accounts. But the original platforms can do things we can't,” Chaudhary explains.

Primary ticketing platforms could notify resale marketplaces when a ticket has been scanned or transferred. They could also reissue tickets to buyers, invalidating the seller's original copy and making double-selling harder.

Purchase limits face a similar challenge. Reticket caps resale prices at the amount originally paid by the seller, but this cannot prevent tickets bought during the original sale from being resold through WhatsApp or other informal channels. Chaudhary argues that linking purchase limits to verified individuals rather than accounts could make bulk buying harder.

Personal AI Agents Enter the Fray

The next challenge may come from AI agents that can search for tickets, compare seating options, and complete purchases on customers' behalf. Platforms will have to distinguish legitimate agent-assisted transactions from automated activity intended to acquire tickets in bulk.

“Soon, many people will simply ask an AI assistant to book tickets for them, and that's completely fair,” Chaudhary says. The question for platforms, he adds, will be whether a genuine fan behind a request and how many tickets they want to purchase.

Gupta similarly argues that platforms should not automatically treat all automation as malicious. “The distinction should be based on identity, authorisation, transaction behaviour, and intent rather than simply whether a browser or an AI agent is involved.”

This could require systems that establish which customer an agent represents, what actions it is permitted to perform, and whether a specific transaction has been authorised. Without those checks, the same tools that help a fan book a ticket could also be used to coordinate purchases across multiple accounts.

來源:Analytics India Magazine · analyticsindiamag.com