The European Union is posing a significant threat to Microsoft, the IT behemoth known for Windows, Office, and the Azure cloud platform, regarding its Bing search engine and related generative artificial intelligence (AI) services. In the event that Microsoft does not reply to a request for information on any threats posed by Bing’s AI capabilities by May 27th, the EU has threatened to punish the company with up to 1% of its annual turnover in the region.
The threat stems from concerns raised by the European Commission under the recently enacted Digital Services Act (DSA). According to a May 17th post on the social media platform X (formerly Twitter), the Commission stated that it wants Microsoft to “provide information under the Digital Services Act on generative AI risks on Bing.”
We compel Microsoft to provide information under the Digital Services Act on generative AI risks on Bing.Bing may pose risks linked to generative AI, such as so-called ‘hallucinations’, deepfakes, as well as the automated manipulation of services that can mislead voters.#DSA
— European Commission (@EU_Commission) May 17, 2024
The Commission mentioned specific possible dangers associated with Bing’s generative AI features, such as “Copilot in Bing” and “Image Creator by Designer.” These dangers include “hallucinations,” in which artificial intelligence creates fiction that is presented as truth; “deepfakes,” which are artificial intelligence-manipulated synthetic media; and automated manipulation of services that might fool voters.
The European Digital Services Act
The Digital Services Act entered into force in November 2022. It is intended to control internet platforms and services. They include search engines, social media, and e-commerce. One of its primary objectives is to address the problems brought forth by artificial intelligence and other new technologies.
Large online platforms must evaluate and reduce any risks connected to their services since then. For example, those involving AI systems in accordance with the DSA. The Act also gives the European Commission the authority to punish non-compliant corporations and demand information from them.
Considering Microsoft’s huge presence in the EU, the possible penalty of 1% of its yearly sales is a substantial amount. Microsoft announced $211 billion in sales for the fiscal year 2023, with a significant amount going toward the EU market.
Should the present market trajectory persist, Microsoft’s earnings for the fiscal year 2024 would surpass $211 billion. A 1% charge in this case may be well over $2 billion, which would be a significant fine for any business, let alone one as big and well-funded as Microsoft.
Microsoft’s Position on AI Ethics
Microsoft has continuously stressed its commitment to responsible AI development and deployment, even though the firm has not yet officially addressed the EU’s particular request addressing Bing AI.
Brad Smith, the President of Microsoft, noted in a recent article the critical role of thoughtful policymaking and democratic processes in developing AI regulations. He stressed that AI regulations should be both effective and well-crafted, focusing on applications that pose the highest risks. Smith argued that these regulations must be outcomes-focused, ensuring they achieve their intended goals. Additionally, he highlighted the importance of creating regulations that can endure the rapid pace of technological advancements and adapt to the evolving expectations of society.
A specialised team and framework for responsible AI have also been formed by Microsoft; these are led by values including inclusivity, responsibility, safety, privacy, and openness. Natasha Crampton, the company’s Chief Responsible AI Officer, has emphasised how crucial it is to include responsible AI concerns across the whole system lifetime, from design to deployment.
The Bing AI Debate and Its Possible Dangers
Since their recent release, Microsoft’s Bing AI features—such as Copilot and Image Creator—have drawn a lot of interest and criticism. Extensive language models and generative AI approaches allow users to produce text, code, and graphics in response to natural language cues.
Although there is much room for improvement in terms of efficiency, creativity, and accessibility with these skills, there are also worries around possible abuse, prejudice, and the dissemination of damaging or false information.
The vulnerabilities connected with generative AI systems are highlighted by the EU’s specific worries about “hallucinations” and deepfakes. These algorithms have the ability to produce extremely realistic but completely fake information, which might be used as a weapon for evil intent like fraud, manipulation, or election meddling if sufficient protections and supervision aren’t in place.
Furthermore, the EU’s reference to the automated manipulation of services may allude to the possibility that AI systems would take advantage of or manipulate online platforms and services, compromising their reliability and integrity.
Finding the Correct Balance: Ethical AI Innovation and Governance in 2024
AI technologies are developing at a rapid rate, which makes responsible innovation and sensible regulation more and more important. While still promoting innovation, the EU has taken a proactive step in addressing the possible hazards linked with generative AI by moving to obtain information from Microsoft under the Digital Services Act.
Microsoft, as well as other major tech companies and pioneers in AI, will have to collaborate closely with legislators and regulators to find the ideal balance between encouraging innovation and averting any harm. This is probably going to require a mix of government supervision, industry best practices, and self-regulation.
Artificial Intelligence Specialist Gabriela Ramos at UNESCO underlined the importance of the moral compass in the area. She made the point that the world is about to change at a rate not witnessed since the invention of the printing press six centuries ago because these all-purpose technologies are revolutionising the way we live, work, and interact. Although AI technology has many applications, Ramos cautioned that in the absence of moral guidelines, it might reinforce prejudice and discrimination in the real world, exacerbate polarisation, and endanger basic freedoms and rights.
The corporation is in line with the goals of the EU under the Digital Services Act with its current efforts in responsible AI research and its support for deliberate, results-focused regulation. Microsoft might be able to overcome this obstacle and keep pushing the limits of generative AI by working with stakeholders and exhibiting a sincere dedication to moral AI practices.
EU inspection of Microsoft’s Bing AI demonstrates the increasing attention being paid to AI governance and legislation on a worldwide scale. Other countries and international organisations are also battling the difficulties presented by sophisticated AI systems, even if the European Union has led the way with the Digital Services Act and the upcoming AI Act.
The Biden administration in the US has started programs to encourage ethical AI development and use, as well as proposing an AI Bill of Rights. The National AI Advisory Committee has stressed how important it is to lower the risks associated with AI systems. At the same time, they don’t forget about innovation and competitiveness.
The Organisation for Economic Cooperation and Development (OECD) has created guidelines for AI. They take into account issues with human oversight, accountability, openness, and fairness. In addition, the UN has created a specific track on AI governance as part of its larger agenda for digital cooperation.
Ensuring the appropriate development and deployment of AI will be imperative in upholding public trust, protecting basic rights, and fully realising the promise of this technology as it becomes more and more ingrained in our everyday lives.
The post EU Warns Microsoft of Potential $2 Billion Fine Under Digital Services Act for Bing AI Risks appeared first on Metaverse Post.