Follow @dorgang

Follow @snbchf

Home › 3) Swiss Markets and News › 3.) Swissinfo Business and Economy › How intelligent are safeguards against AI abuse?

Previous post Next post

How intelligent are safeguards against AI abuse?

Published on September 10, 2024

See on Internet Archive

by
Swissinfo
My articles My site About me My videos My books
Follow on:Twitter Facebook Google +

How intelligent are safeguards against AI abuse? — Artificial intelligence is expected to have a big impact on people's lives. Keystone

Generated with artificial intelligence.

Artificial intelligence (AI) holds the power to either enhance or damage people’s lives. The global community, including Switzerland, is attempting to thrash out a set of rules to ensure AI develops into a friend rather than a foe.

This content was published on September 10, 2024 - 14:07

When not covering fintech, cryptocurrencies, blockchain, banks and trade, swissinfo.ch's business correspondent can be found playing cricket on various grounds in Switzerland - including the frozen lake of St Moritz. 

+ Get the most important news from Switzerland in your inbox

The Council of Europe, the continent’s guardian of human rights and democracy, has drawn up a treaty (Convention) that aims to steer AI along a path for good.

Switzerland endorses the Convention, having played a role in drafting the text. But the Alpine state was not among the first set of countries to sign a binding commitment to uphold its values.

SWI swissinfo.ch looks at the Convention’s content and Switzerland’s contribution to controlling AI.

What is the treaty?

The document presented to the world in May has a rather long name: The Council of Europe Framework Convention on Artificial Intelligence and Human Rights, Democracy and the Rule of Law.

+ Is the Swiss negotiator a US stooge?

The Convention sets out to preserve the “protection and promotion of human rights, democracy and the rule of law” when applying AI to a range of use cases.

Signatories pledge to amend existing laws or create legislation to incorporate the specific challenges posed by AI. Potential pitfalls of AI include spreading misinformation or corrupting AI systems with bad data that can result in biased decision and discrimination.

At the same time, the Convention stipulates that regulations should encourage and support the proper applications of AI, which could bring many benefits.

The Council of Europe has 46 member states, but other countries, such as the United States and Japan, were invited to help draft the Convention to boost its international credentials.

On September 5, the Convention collected its first ten signatories, including the European Union, the US, Britain and Israel. You can find a continuously updated list of signatories hereExternal link.

Why has Switzerland not signed?

The Swiss Federal Office of Communications is currently drawing up recommendations for how Switzerland could regulate AI in future. These recommendations will be presented to the government by the end of the year.

+ Switzerland ‘could do more on AI governance’

The normal process then includes a consultation period with representatives from the economy, civil society, the cantons and other interested parties before parliament debates the issue.

Any proposed legislative changes could also be challenged by citizens, resulting in a potential nationwide vote. Signing the Convention is not an option until at least after the communications ministry presents the findings of its AI regulation review.

The ministry told SWI swissinfo.ch that it welcomed the initial signatures to the Council of Europe Convention.

“Whether the Convention will form the basis for Switzerland’s own national AI policy, and the timeframe in which a possible signature and ratification of the Convention would be sought, is currently being examined as part of the review,” the ministry said.

The Convention has, however, already been endorsed by the government, which signalled its approval of the document in May.

In addition, Swiss diplomat Thomas Schneider, who heads the Council of Europe’s Committee on AI, led negotiations between countries as the Convention was being drafted.

What is the reaction to the Convention?

Given the range of countries involved in the negotiating rounds, compromises were always inevitable.

All the countries involved were content with the final outcome of the talks, but some members of civil society have picked up on the compromises in a negative light.

NGOs, private companies and other non-state actors were invited to offer their views but did not have seats at the negotiating table. “This was a bad start,” according to the US Center for AI and Digital Policy thinktank.

The ethics of artificial intelligence

The challenges are immense for Switzerland, which is one of the leading developers of artificial intelligence (AI).

Another contentious point is that the Convention excludes private companies from its scope and instead focuses on state activities.

The NGO AlgorithmWatch Switzerland sounded an alarm about this when the talks were well underway in March. “The first international rulebook on AI could thus give corporations a free pass to develop and use AI according to their own interests,” said AlgorithmWatch Switzerland Executive Director Angela Müller at the time. “The negotiating states must ensure that AI serves the interests of humanity and not those of a few big corporations.”

Another criticism from NGOs is that the Convention lacks enough legal bite to force signatory states to comply with its demands.

What are the pros and cons of AI?

AI can be applied to practically any walk of life. The Explanatory Report of the Convention lists the following use cases: “the distribution of social welfare benefits, decisions on the creditworthiness of potential clients, staff recruitment and retention processes, criminal justice procedures, immigration, asylum procedures and border control, policing, and targeted advertising”.

Is AI going to help or hurt us?

Computers are proving capable of performing tasks that require human intelligence and to influence our decisions. Should we be letting them?

Mar 14, 2022

View the discussion

Machines can be taught to quickly filter through vast troves of data to piece together relevant information to achieve certain tasks. A new form of AI, generative AI, is designed to “think for itself” to come up with original insights and solutions.

Regulations are seeking to find a balance between allowing AI to be creative and generate answers faster than humans, while curbing its potential to churn out anti-social content.

Is there any other AI regulation?

The Council of Europe Convention is one of many global and more focused attempts to control AI.

Last year the European Union released its AI Act to regulate the use of the technology.

In November 2023, 28 countries, including the US, China, the EU, Australia and Britain, signed the Bletchley Declaration to work together on AI safety research.

The Council of Europe Convention sets out principles but recognises that individual countries will implement them in different ways, taking into account varying national interests on matters such as border control and policing.

Switzerland is one of many countries currently working out the best AI regulatory fit for its national interests.

Edited by Marc Leutenegger/ts

Full story here Are you the author?

About Swissinfo

SWI swissinfo.ch – the international service of the Swiss Broadcasting Corporation (SBC). Since 1999, swissinfo.ch has fulfilled the federal government’s mandate to distribute information about Switzerland internationally, supplementing the online offerings of the radio and television stations of the SBC. Today, the international service is directed above all at an international audience interested in Switzerland, as well as at Swiss citizens living abroad.

My articles My site About me My books
Follow on:Twitter Facebook Google +