/industry-wired/media/agency_attachments/2024/12/04/2024-12-04t130344212z-iw-new.png)
/industry-wired/media/agency_attachments/2024/12/04/2024-12-04t130332454z-iw-new.jpg)
/industry-wired/media/post_attachments/wp-content/uploads/2023/12/The-Worlds-First-AI-Law-EUs-Goal-and-Challenges.jpg)
EU is working towards a consensus on the first global law to regulate AI
The European Union (EU) is attempting to reach an agreement on comprehensive laws to govern artificial intelligence (AI) to defend the rights, privacy, and data protection of EU people. The proposed AI regulation in the EU seeks to regulate systems depending on the amount of danger they pose. The discussions have centred on how to govern generative AI, such as ChatGPT and Google's Bard chatbot. Some member states are afraid that excessive regulation would impede innovation and prevent European AI behemoths from emerging.
On Wednesday, the European Union will attempt to reach an agreement on broad regulations to oversee artificial intelligence, following months of contentious discussions, particularly on how to supervise generative AI like ChatGPT.
After the ChatGPT bot broke onto the scene last year, showing AI's rapid achievements, the EU is hurrying to establish the world's first comprehensive AI regulation.
ChatGPT dazzled with its capacity to generate poetry and essays in seconds.
Proponents of AI argue that it will help mankind by revolutionising everything from labour to healthcare, while others are concerned about the threats it poses to civilization, worrying that it would plunge the world into unparalleled turmoil.
Brussels is determined to bring big tech to heel by arming itself with a formidable legal arsenal to safeguard EU citizens' rights, particularly those concerning privacy and data protection.
In 2021, the European Commission, the EU's executive arm, proposed an AI law that would govern systems depending on their level of danger. For example, the bigger the risk to citizens' rights or health, the larger the responsibility of the systems.
Negotiations on the final legislative wording began in June, but a heated dispute in recent weeks about how to handle general-purpose AI such as ChatGPT and Google's Bard chatbot threatened to derail negotiations at the last minute.
Some member states are concerned that excessive regulation would discourage innovation and harm the possibilities of establishing European AI behemoths to compete with those in the US, notably ChatGPT's founder OpenAI and tech heavyweights like Google and Meta.
Negotiators from the European Parliament and EU member states will meet on Wednesday at 1400 GMT, and discussions are anticipated to extend until the evening.
Although no firm timeline has been set, key EU officials have stated repeatedly that the law must be finalized by the end of 2023.
Chasing Local Champions
EU ambassadors, industry insiders, and other EU officials have cautioned that the discussions might finish without a deal since critical problems remain unresolved.
Others have argued that even if a political consensus is reached, numerous sessions would be required to iron out the technical elements of the bill.
Even if EU negotiators reach an agreement, the bill would not go into effect until at least 2026.
The key area of contention is how to control so-called foundation models, which are meant to fulfil a range of jobs, with France, Germany, and Italy advocating for their exclusion.
"France, Italy, and Germany don't want a regulation for these models," said German MEP Axel Voss, a member of the special parliamentary committee on AI.
The parliament, on the other hand, feels it is "necessary... for transparency" to regulate such models, according to Voss.
Late this month, the three largest EU economies issued a document advocating for an "innovation-friendly" approach to the AI Act.
Berlin, Paris, and Rome oppose the inclusion of restrictive requirements for foundation models in the statute, instead arguing that they should follow standards of behaviour.
Many believe this shift in perspective is motivated by a desire to avoid impeding the growth of European champions – and maybe to assist businesses like France's Mistral AI and Germany's Aleph Alpha.
'Significant progress'
Another source of contention is remote biometric surveillance, which involves facial recognition using camera data in public settings.
The EU parliament seeks to outright ban "real-time" remote biometric identification devices, which member countries reject. Initially, the panel advocated exclusions to discover possible victims of crime, such as missing children.
There have been suggestions that MEPs may give ground on this issue in exchange for concessions elsewhere.
Thierry Breton, the EU's chief tech enforcer, is optimistic about a deal.
"Significant progress has been made by both sides," said Breton, the EU's internal market commissioner, on Tuesday, referring to the mandate granted to Spain, which now holds the rotating EU presidency, to finish talks on Wednesday.
Concerns about the effect of AI and the necessity to monitor the technology are shared globally.
To limit the hazards of AI, US President Joe Biden signed an executive order in October regulating the technology.