Italy's stop to ChatGPT is BS
As a teenager, I soon had to come to terms with a rough truth: Italy is seldom first. It's highly unusual to see the name of my country connected to profitable companies, technological breakthroughs, or at least pop culture mainstays. The grandeur of the Roman Empire and Renaissance city-states is long gone and even the Italian economic miracle of the 60s, an industrial and social revolution that gave Italy a respectable seat in the G8, is now a distant memory.
I can barely believe that just 30 years ago Italy was the 4th biggest global economy, ahead of ex-colonial powers such as the United Kingdom or France. Sure, we still manage to retain some face by exporting our beloved cultural icons abroad: fast cars, fashion, and pretty good food. However, this also connects us to our heritage in a grotesque way, as I found out the last time I took a taxi ride abroad.
«Oh, you're from Italy» the driver told me with a big smile «You are one of the pasta people then!»
You may now understand why, as a citizen of a country whose relevance is waning on the global scene, I'm not accustomed to seeing Italy at the center of international interest. Imagine my utter disbelief to see my home country all over the news, while lazily strolling in Cape Town International Airport, on the other side of the world.
"Italy's first West country to ban ChatGPT". Now, this is some real leadership! Our Roman ancestors would be so proud of us.
Think Of The Children!
From a formal standpoint, I have no reason to assume the sanction against OpenAI is unlawful - neither I have the legal training required to make a similar assessment. The measure was issued on the 31st of March by the Italian Garante, which is a government agency established in 1996 with the mandate to safeguard the data privacy of Italian citizens. The Garante intimated OpenAI to stop gathering data from users in Italy, on the grounds of various alleged privacy laws infractions, and gave a 20-day ultimatum to the US company before starting to issue mighty fines of up to 4% of the annual turnover or €20 million (whichever is higher), which is in line to the European GDPR regulation. However, the official statement from the Garante and the context of the ruling contain many troubling details that don't sit right with me.
Why, of all nations, Italy is the first (and at the time of the writing, the only one) to issue such a ban? Let's be brutally honest here: the percentage of people fluent in English in this country is one of the lowest in Europe. On top of that, generally speaking, Italian society has quite a conservative approach toward innovation and technology, since the fundamental fabric of its economy is characterized by family-led SMEs, often operated over the course of multiple generations.
All things considered, how many people are really taking advantage of a tool that was not trained in the Italian language, and to what extent? How much data was really stolen from unaware Italian cybernauts? I can only assume that the answers to these questions are incredibly underwhelming. On the other hand, one could say that numbers don't matter when our privacy is at stake: dura lex sed lex. OpenAI is failing to inform and to protect its users, so it will be punished.
However, what is really bewildering is a further claim from the Garante. Apparently, OpenAI is not correctly verifying the age of children under the age of 13, corrupting our youth and exposing kids to a dangerous AI chatbot. To add some context to international readers, children's rights - and the need of protecting them - is a recurring topic in the current Italian debate, often in relationship with the alleged dangers coming from the LGBTQ+ community.
I would never have imagined that OpenAI would be included in the blacklist of child offenders, along Gay Prides and gestational surrogacy. The most surprising fact is that the Garante worries about regulating children's access to the Internet only now, in 2023, after almost 3 decades of idleness, and it reacted by immediately showing its muscles. I wonder if the officials working at the Garante were imprisoned in a cryo-chamber during WW2, and after being exhumed last Friday they panicked at the state of our current world and hit the first big red button at hand's reach labeled "STOP!". If that's the case, they have many buttons to push now.
The third and final claim is equally absurd. Apparently, ChatGPT can be a dangerous aid in fabricating and distributing fake news. I would like to report to the Garante that Social Media (and by extension, search engines) are full of fake news, conspiracy theories, and intrusive ads featuring obvious scams. I could create a WhatsApp group chat right now with the only goal to share a dubious article about how pineapple pizza is a covert medium for 5G microchips. Where is Meta's ban? Honestly, even some reputable media outlets are getting lax with their story verification. Why the Garante is striking the new kid on the block, and why only now?
Conclusion
I can't dispute the decision from the Garante. OpenAI is failing to respect privacy laws and was not timely in reacting to a serious data breach on the 20th of March: it is difficult to speculate was kind of legal countermeasures the company will take now as a remediation. However, the decision from OpenAI to block ChatGPT in Italy makes perfect sense and it has its merits as a possible long-term measure: why even worry about a country that barely brings in any revenue?
Still, the unusually prompt and firm decision from the Garante has the bitter aftertaste of political clout used to hide fear and ignorance: let's be the first to act, so we can continue to be last.