Factbox: Governments race to regulate AI tools

Sept 11 (Reuters) – Rapid advances in artificial intelligence (AI), such as Microsoft-backed OpenAI’s ChatGPT, are complicating governments’ efforts to pass laws governing the technology’s use.

Here are the latest steps national and international governing bodies are taking to regulate AI tools:

Australia

* Planning Rules

Australia will build search engines like Alphabet’s Google and Microsoft’s Bing, draft new codes to prevent the sharing of AI-generated child sexual abuse content and the production of deepfake versions of the same content, the country’s internet regulator said on September 8.

Britain

* Planning Rules

The Financial Conduct Authority, one of several state regulators tasked with creating new guidelines to cover AI, is consulting with the Alan Turing Institute and other legal and academic institutions to improve understanding of the technology, a spokeswoman told Reuters.

Britain’s competition regulator said in May it would begin examining the impact of AI on consumers, businesses and the economy, and whether new controls are needed.

China

* Provisional rules applied

China has issued a set of temporary measures effective August 15 to manage the generative AI industry, requiring service providers to submit security assessments and obtain approval before releasing mass-market AI products.

Following government approval, four Chinese technology companies including Baidu Inc ( 9888.HK ) and SenseTime Group ( 0200.HK ) launched their AI chatbots to the public on Aug. 31.

European Union

* Planning Rules

EU lawmakers agreed to changes to the bloc’s draft AI law in June. Lawmakers will now have to work out the details with EU countries before the draft rules become law.

The biggest issue is expected to be facial recognition and biometric surveillance where some MPs want a total ban while EU countries want exceptions for national security, defense and military purposes.

France

* Investigating potential violations

France’s privacy watchdog CNIL said in April it was investigating several complaints about ChatGPT after the chatbot was temporarily banned in Italy for suspected violations of privacy rules.

France’s National Assembly approved the use of AI video surveillance during the 2024 Paris Olympics in March, ignoring warnings from civil rights groups.

G7

* Looking for input on rules

A meeting of Group of Seven (G7) leaders in Hiroshima, Japan, in May acknowledged the need for governance of AI and immersive technology, and ministers agreed to discuss the technology as the “Hiroshima AI Process” and report on the results by the end of 2023.

G7 nations should adopt “risk-based” regulation on AI, G7 digital ministers said after a meeting in April.

Ireland

* Looking for input on rules

Generative AI needs to be regulated, but governing bodies must figure out how to do it properly before rushing into restrictions that “won’t really stand up”, Ireland’s data protection chief said in April.

Israel

* Looking for input on rules

Israel Innovation Authority national AI planning director Ziv Katzir said in June that Israel has been working on AI regulations “for the past 18 months” to strike the right balance between innovation and the protection of human rights and civilian security. .

Israel published a 115-page draft AI policy in October and is gathering public feedback before a final decision.

Italy

* Investigating potential violations

Italy’s data protection authority plans to review other artificial intelligence platforms and hire AI experts, a top official said in May.

ChatGPT became available again to users in Italy in April after being temporarily banned in March due to concerns by the National Data Protection Authority.

Japan

* Investigating potential violations

Japan expects to introduce rules by the end of 2023 that are closer to the U.S. stance than the stricter ones planned in the European Union, an official close to the deliberations said in July, as it looks to the technology to spur economic growth from its leader in advanced chips.

The country’s privacy watchdog said in June that it had warned OpenAI not to collect sensitive data without people’s permission and to minimize the sensitive data collected.

Spain

* Investigating potential violations

Spain’s data protection agency said in April that it was launching a preliminary investigation into a possible data breach by ChatGPT. It has asked the EU’s privacy watchdog to assess the privacy issues surrounding ChatGPT.

United Nations

* Planning Rules

The UN Security Council held its first formal discussion on AI in New York in July. The council addressed the military and non-military applications of AI, which “could have very serious implications for global peace and security”, UN Secretary-General Antonio Guterres said.

Guterres in June supported a proposal by some AI officials to create an AI watchdog like the International Atomic Energy Agency, but noted that “only member states can create it, not the UN Secretariat.”

The UN Secretary-General has announced plans to launch a high-level AI advisory body by the end of the year to regularly review AI governance arrangements and offer recommendations.

US

* Looking for input on rules

Congress will hold three hearings on AI on September 11, 12 and 13 as it works on legislation to mitigate the risks of the emerging technology. Discussions include AI forums and meetings with various members, including Microsoft ( MSFT.O ) President Brad Smith and Nvidia ( NVDA.O ) Chief Scientist William Daley, Meta Platforms ( META.O ) CEO Mark Zuckerberg and Tesla CEO Elon Musk. House and Senate Subcommittees.

Washington DC District Judge Beryl Howell ruled on August 21 that artwork created by AI without any human input cannot be copyrighted under US law, after the Copyright Office rejected an application filed by computer scientist Stephen Thaler on behalf of his DABUS system. .

The US Federal Trade Commission (FTC) opened a wide-ranging investigation into OpenAI in July over claims it violated consumer protection laws by putting personal reputations and data at risk.

Generative AI raises competition concerns and is the focus of technology bureaus, including the FTC’s Office of Technology, the agency said in a blog post in June.

Senator Michael Bennett wrote to leading tech companies in June urging them to label AI-generated content and limit the spread of content intended to mislead users. In April, he introduced a bill to create a task force to look at US policies on AI.

Our Standards: The Thomson Reuters Trust Principles.