© IE Online Media Services Pvt Ltd
Tags:
Journalism of Courage
Italy late last month banned ChatGPT due to concerns over the collection of personal data and the lack of guardrails to prevent minors from accessing the AI chatbot. The Italian Data Protection Authority said it had imposed a temporary but immediate restriction on the company behind ChatGPT, OpenAI, to stop them from processing the data of Italian users. The privacy watchdog also plans to launch an investigation into whether the chatbot breaches the EU’s General Data Protection Regulation (GDPR).
The EU is often at the forefront of tech regulation, and is perhaps the biggest reason behind Apple possibly opting for USB-C with the iPhone 15. The 27-nation union is expected to take a pretty restrictive stance on AI and has already proposed legislation known as the European AI Act.
The European AI Act aims to introduce a common regulatory and legal framework for artificial intelligence in the EU, covering all sectors except for the military and all types of artificial intelligence. It will classify different AI tools according to their perceived level of risk, from low to unacceptable, and impose different obligations and transparency requirements on those who provide or use them. The AI Act will also work in tandem with other laws such as the General Data Protection Regulation (GDPR).
However, when the act was first drafted, officials hadn’t accounted for the rapid pace at which AI would advance in 2022 and beyond, and turn capable of generating content and art comparable to humans.
A Reuters report showed that the EU’s draft rules may include ChatGPT under the “General Purpose AI Systems” (GPAIS) category, which describes tools that can be adapted to perform a number of functions. It remains to be seen if GPAIS will be deemed high-risk.
Regardless, Italy’s move to outright ban ChatGPT has inspired other European countries to study if harsh measures are necessary to control chatbots like it and if they should coordinate such actions.
Last week, the UK announced plans for regulating AI, outlining an all-encompassing approach for regulating the technology that has reached unholy levels of hype. But rather than establishing new regulations like the EU, the UK government is asking regulators in different sectors to apply existing regulations to AI.
In a white paper published last week, the Department for Science, Innovation, and Technology (DSIT), outlines 5 principles it wants companies to follow – safety, security and robustness; transparency and explainability; fairness; accountability and governance; and contestability and redress.
At this stage, the UK is not proposing restrictions on ChatGPT or any other kind of AI. Instead, the country looks to be settling for a rather light-touch approach. The DSIT says that legislation “could” be introduced “to ensure regulators consider the principles consistently,” but did not specify exactly when.
“Over the next twelve months, regulators will issue practical guidance to organisations, as well as other tools and resources like risk assessment templates, to set out how to implement these principles in their sectors,” the government said.
The NITI Aayog, the planning commission of India, has issued some guiding documents on AI, such as the National Strategy for Artificial Intelligence and the Responsible AI for All report. These documents outline the vision, goals, and principles for developing and deploying AI in India, with an emphasis on social and economic inclusion, innovation, and trustworthiness.
However, these documents are not legally binding and do not address some of the key issues and challenges related to AI, such as accountability, liability, transparency, explainability, and human oversight.
The United States doesn’t have comprehensive federal legislation on AI yet. It instead has a patchwork of various current and proposed AI regulatory frameworks, focused on specific AI-use cases such as AI in recruitment or employment. The U.S. also has a new Trade and Technology Council (TTC) with the EU, which aims to align on some common principles and goals for AI governance.
However, the country’s National Institute of Science and Technology (NIST) does promote an AI Risk Management Framework framework that gives companies using AI guidance on how to “improve the ability to incorporate trustworthiness considerations into the design, development, use, and evaluation of AI products, services, and systems.”
But since the framework is on a voluntary basis, companies that do not implement it will not be penalised in any way. The US also so far seems to be taking no action to limit ChatGPT in the country.
Last month, an AI think tank filed a complaint with the FTC in a bid to stop OpenAI from any further commercial deployments of GPT-4. The Center for Artificial Intelligence and Digital Policy (CAIDP) claims OpenAI has violated a section of the FTC Act, alleging the AI company of deceptive and unfair practices. The complaint could lead to an investigation into OpenAI and suspension of commercial deployment of its LLMs.
While China hasn’t officially blocked ChatGPT, OpenAI does not allow users to sign up for the chatbot in the country. OpenAI also blocks users from other countries with heavy internet censorship such as Russia, China, North Korea, Egypt, Iran, Ukraine, and a few more.
However, despite ChatGPT’s unavailability, China continues to be a global leader in AI and it’s been investing heavily in improving AI research. Multiple tech companies in China are developing LLMs (Large Language Models) of their own, including search engine giant Baidu — and they’ve already announced ChatGPT rivals.
