Yahoo News Why the Willy Wonka shambles shows how dangerous AI can be
The fallout from the disastrous 'experience' in Glasgow has amused many but raises some serious questions.
The disastrous Willy Wonka experience left children upset, parents fuming and the internet a bit amused. Now, an expert in artificial intelligence has warned the incident raises questions of whether we're at a tipping point for AI scams.
Parents paid £35 per ticket for the 'shambolic' Glasgow event, centred around the Roald Dahl classic Charlie and the Chocolate Factory. Children arrived dressed up on 24 February, but were given an 'experience' alarmingly different to Willy Wonka's fantasy world.
Read more: The fallout to the disastrous Willy Wonka Experience, explained (Yahoo News)
The website promoting the event featured elaborate art and extensive promotional text but many have speculated at least some of the promotional content was generated by AI. Actors who worked on the event also complained of the low quality of the scripts they were given and said they looked like it was written by AI.
While the promoters have promised to issue full refunds, David Emm, principal security researcher at Kaspersky, told Yahoo News UK that the incident highlighted the explosion of AI tools in recent years - and raises questions over whether AI will lead to more lucrative opportunities for some to set up "cheap and easy" ways to deceive people.
He said that scammers throughout history have been among the first to leap on new opportunities to exploit other people. "With any new type of technology, you get the sort of early adoption then it suddenly hits a tipping point where it suddenly becomes widespread and popular and that's kind of what's happened here with AI," he said.
AI makes scams easier to produce
Emm pointed out that "scammers capitalise on human fallibility and what AI does is give them a bit of an easier way to do it". He noted when people think of the risks of AI they often imagine deep fakes and sophisticated scams. Emm said these still require "significant investment" and scams offering experiences of lucrative events to the public can be "quick and easy" with tools like ChatGPT.
Emm highlighted an event in 2008 that offered a "winter wonderland" in the New Forest with "Hollywood special effects." The event told 40,000 tickets but when customers arrived they found reindeer that looked unwell, "chained up" huskies and the whole area covered in mud.
Trading Standards received 5,000 complaints about the event and two men were prosecuted for misleading customers. Emm said what the Willy Wonka event instead sparks debate about the "shrewd" use of AI. He added: "The ability to generate convincing - yet entirely fictional - visuals and narratives can be exploited with unprecedented sophistication."
He said: "I think what AI lends to this is the ability to easily and cheaply to produce content of that sort which in this case inflates what was actually on offer. I know some of the actors involved in this were talking about how rubbish the sort of AI-based scripts were and that highlights the fact that people really didn't put much effort into developing that."
Read more: Oompa Loompa from Glasgow Wonka experience breaks silence (The Independent)
The Willy Wonka organiser has since apologised, posting on Facebook: "I am truly sorry for any upset and disappointment caused at the weekend. Refunds have been issued and will continue to do so. This was an event gone wrong, The House of Illuminati will NOT be holding any other events in the foreseeable future."
Emm also said another big area scammers will be using AI is phishing emails. He said AI tools "make crafting phishing emails easier and in a way more credible because of fewer grammar errors and they can also produce spectacular graphics to go with something."
He added: "At a level of very mundane scams I think we will see more of it."
How to avoid AI scams
Not only is the likelihood of AI scams increasing as tools like chatbots become more prevalent, society is constantly moving more and more online. Emm pointed out that COVID-19 pushed elderly people who had never engaged with some technology before to use it to engage with their families, leading to ever more people being exposed to scams who might not have the tools to identify them.
Emm told Yahoo News UK there are several ways to identify events that might not be what they seem. He pointed out how, in the Willy Wonka example, the company that organised the event pointed out they had no connection to the owner of the rights to the character. The latest film about the famous chocolate maker was made by Universal Studios.
It is also worth inspecting thoroughly any graphics and images. Although AI art is getting more and more sophisticated it struggles with producing hands, making the lighting look realistic and keeping everything uniform. AI art also struggles with words in the images, in the image produced to promote the Willy Wonka experience the advertising text was littered with typos.
Emm also said that if something looks "too good to be true, it probably is" and one way to check is to call the people behind whatever is being promoted and get them to answer in person what is on offer. AI may be used to lure people in but, at the end of the day, a real person is behind them. If no phone number is available then alarm bells should be ringing.
Ultimately Emm pointed out: "It belongs to all of us to be a bit more suspicious about what is out there."
