AI News, Translation: Chinese AI Alliance Drafts Self artificial intelligence
- On 24. juli 2019
- By Read More
Regulation of Artificial Intelligence: Europe and Central Asia
Back to Index of Regulation of Artificial Intelligence On May 25, 2018, the EU’s General Data Protection Regulation (GDPR) took effect. It applies directly in all EU Member States without any implementing legislation needed. Among other rights, the GDPR guarantees individuals the right to have a decision based solely on automated processing (an algorithm) be made or reviewed by a natural person instead of a computer. On February 16, 2017, the European Parliament adopted a legislative initiative resolution in which it recommended a range of legislative and nonlegislative initiatives in the field of robotics and AI to the European Commission. Among other things, it called on the European Commission to adopt a proposal for a legislative instrument providing civil law rules on the liability of robots and AI, “to propose common Union definitions of cyber physical systems, autonomous systems, smart autonomous robots and their subcategories,”
(Communication) in which it outlined the EU’s approach to taking advantage of the opportunities offered by AI and addressing the challenges posed by it. From 2014–2017, the EU invested €1.1 billion (about US$1.26 billion) in AI-related research and innovation under the Horizon 2020 research and innovation program. Among other things, the Communication emphasizes that the EU has to ensure that no one is left behind in the digital transformation and that AI is developed and applied based on the EU’s values and fundamental rights. It is also reviewing existing rules on safety and civil law liability. The Commission published a follow- up communication and a coordinated plan that built on the original Communication in December 2018. On December 18, 2018, the High-Level Expert Group on Artificial Intelligence (AI HLEG)— fifty-two experts from academia, civil society, and industry appointed by the Commission— released draft AI ethics guidelines, which set out a framework for designing trustworthy AI. According to the guidelines, trustworthy AI must “respect fundamental rights, applicable regulation and core principles and values, ensuring an “ethical purpose,”
The ESAs plan to carry out a new monitoring exercise if development of the market and market risks warrant it. In 2016, the European Commission published a communication on intelligent transport systems. It proposed, inter alia, the adoption of a legal framework at the EU level by 2018 developed through learning by experience. In 2018, the Commission published another communication that outlined the EU strategy to enable the EU and its Member States to “be a leader in delivering safe, efficient, socially responsible and environmentally friendly driverless mobility for EU citizens.” The 2018 Communication is accompanied by two legislative proposals to implement that objective, one for a revision of the General Safety Regulation, and one for a proposal for amendments to the directive on road infrastructure safety. With regard to intelligent transport systems, the Commission published a decision in December 2018 which sets out an updated working program to achieve the objectives of the 2010 Directive on Intelligent Transport Systems. In the August 2018 meeting of the GGE LAWS, the EU stated that it “considers it worthwhile to review more regularly and systematically the fast-paced developments in the area of emerging technologies, including artificial intelligence (AI), providing an opportunity to technical experts to share information on autonomous technologies relevant for our work.” The EU emphasized that it is necessary that “humans remain in control of the development, deployment and use with regard to possible military applications of emerging technologies, including AI, and prevent the creation and use of harmful applications.” Back to Top The Austrian government in its 2017–2022 government program stated that “new digital technologies like AI, robotics, and blockchain will have unforeseeable effects on our society.” It envisages the use of AI in public administration and therefore wants to establish a legal framework in order to “use innovative, new business models and technologies for the development of society while simultaneously safeguarding the data autonomy/sovereignty of citizens and consumers.” In 2017, the Austrian Federal Ministry for Transport, Innovation, and Technology (BMVIT) established the Austrian Council for Robotics and Artificial Intelligence (Österreichischer Rat für Robotik und künstliche Intelligenz). The Council consists of nine experts from research, academia, and business.
However, the law provides that the driver must remain responsible to take over the driving functions if necessary. Autonomous vehicles may only be tested on the road if they have car insurance and transmit certain data to the BMVIT. Test drives on public roads may only be conducted after the vehicles have been tested sufficiently in advance. The results of the tests must be transmitted to the BMVIT after the end of the trial period. Information on critical situations, accidents, and their causes must be reported without undue delay. In addition, all test vehicles with automated or networked systems must be equipped with an accident data recorder that must be activated during all test drives. At the April 2018 meeting of the GGE of the High Contracting Parties to the Convention on Conventional Weapons (CCW), Austria stated it believes that LAWS require a clear regulation and that it is in favor of “establishing a legally binding instrument in order to prohibit lethal autonomous weapon systems that are not under meaningful human control.” Austria emphasized that “meaningful human control over critical functions”
Increased IT-security for business Denmark invested DKK 750 million (about US$115 million) in 2018, and will invest DKK 125 million (about US$19 million) annually until 2025 to fund initiatives connected to these goals. The initial sum includes an extra start-up amount, aimed at establishing a number of new initiatives. Possible initiatives include changed tax provisions to allow for greater deductions for investment in technology. >Examples of changes to regulations include rules that enable new business models. According to the government the purpose behind the new business models is to test new areas, as is done today in the fields of driverless transport and within FinTech. The success of the initiatives also relies on other factors such as increased spending on small businesses in general, enabled by the guarantee agreement with the European Investment Fund and Vækstfonden (Danish Growth Fund) of DKK 1.6 million (about US$245,000). In addition Denmark is increasing its spending on research.
In 2018 it will spend DKK 80 million (about US$12 million) on “digital technologies such as artificial intelligence (AI), big data, Internet of Things, IT-security, etc.” Danish efforts to use AI in the welfare sector—for example, in targeting child welfare benefits fraud—has been criticized as potentially threatening to privacy and other important values. The Nordic countries have also joined together in a number of collaborative responses to AI, including together with the Baltic States. Danish law provides for testing of driverless cars. The current trial term is from 2017 to 2020. The Vejdirektoratet (Danish Road Directorate) has issued a binding Guide on driverless cars. Applications for testing are determined by the Færdselsstyrelsen (Danish Road Safety Agency). Denmark has also tested driverless buses. Denmark has tested autonomous vehicles in perfect weather and in rainy weather. While the results were positive for operation in clear weather, the autonomous vehicles tested in rainy weather had a success rate of 69–92%. A majority of Danes believe autonomous vehicles will be driving on the roads by 2030. Denmark has not passed any legislation banning LAWS.
The two principles are that of loyalty (loyauté, which can also be translated as “faithfulness”) and vigilance/reflexivity. The first principle refers to the idea that algorithms should not betray the interests of its users—not just as consumers but more broadly as citizens and members of communities or other groups whose interests may be affected by the algorithm. The second principle refers to the idea that the constant evolution and unpredictability of AI requires methodical, deliberative, and regular re-examination by all stakeholders. The CNIL’s six concrete recommendations include the following: In 2017, the government established a task force to develop strategies to boost France’s leadership in the field of AI, including two working groups on how to manage the economic, social, ethical, national sovereignty, and national security implications of AI. The resulting report covered several themes, including the importance of AI serving “autonomous individual development, necessary for a functioning democratic society,”
test environment, which includes a self-driving vehicle test track in Zalaegerszeg, in western Hungary. Back to Top The Icelandic Prime Minister addressed the issue of AI at the Global Positive Forum in 2018, stating that it is important that AI serve all and not just the few. Iceland is also the home to an independent institute that focuses on AI, the Icelandic Institute for Intelligent Machines (IIIM/Vitvélastofnun Íslands). The Institute has issued a recommended ethics policy to be used when conducting research and development on AI. As noted above, the Nordic countries have also joined together in a number of collaborative responses to AI, including together with the Baltic States. In 2016, the Icelandic Road and Coastal Administration (Vegagerðin) issued a report on self- driving cars in Iceland. It concluded that the greatest variable risk in connection with self- driving cars in Iceland is the Icelandic weather. At the time of publication there had been no convincing technological solutions to the risks associated with the weather. Iceland’s readiness to adopt autonomous vehicles has also been reviewed by the European Commission, which concluded that Icelandic weather was a problem and that the Icelandic infrastructure might need to be altered to accommodate autonomous vehicles. It was further suggested that autonomous testing by Volvo conducted in Gothenburg should be followed closely. Iceland does not have an active military.
In this context, AI experts in Italy have highlighted the need for the country to deepen its commitment to pan-European AI initiatives by creating or diversifying government, social, and investment programs that would allow access to the larger EU market for Italy-based AI projects. Also, experts note that Italy should put into place stronger conditions to attract foreign AI researchers and entrepreneurs. In July 2018, a group of higher education institutions created the Artificial Intelligence and Intelligent Systems Lab, with the aim of strengthening the country’s “basic and applied research in AI, support the country’s ICT industry by promoting technology transfer from research to entrepreneurship, and promote the adoption of AI solutions in the public administration.” In addition, the Italian Association for Artificial Intelligence, established in 1988, has about nine hundred members from academia, the government, and private companies, and holds an annual conference. The Association’s areas of work are knowledge representation and reasoning, machine learning, deep learning, optimization, planning, ontologies, robotics, natural language processing, and multi-agent systems. In particular, the Association has favored projects aimed at In March 2018, the Agency for Digital Italy published a White Book on Artificial Intelligence that describes the current status of AI in Italy, calls on all stakeholders to improve access to AI in Europe and Italy, proposes a new common culture for innovation in public services, and sets forth challenges related to AI for the Three-Year Plan for Information Technology in the Public Administration, published in 2017. These challenges include In particular, the White Book states the government’s interest in facilitating the adoption of AI strategies and approaches at public agencies in order to provide the general public with faster and more efficient services in, among others, the areas of public health, education, the judicial system, and public security. The White Book recognizes the need to update the legal and regulatory framework for AI in Italy, balancing public and private interests, as well as abiding by the principle of transparency of administrative acts, protection of privacy and copyright, fostering accountability, and establishing an effective system of social control for the protection of the fundamental rights of the person. The White Book includes ten recommendations for consideration by the government, including the creation of a “National Competence Center and a Trans-disciplinary Center on AI,”
focusing on the following objectives: The call for expressions of interest in becoming a member of the Group states that the National Strategy will be the subject of public consultation. Concerning self-driving cars in Italy, a Ministerial Decree of 2018 provides that the experimentation or testing of autonomous vehicles on a protected site or public streets must be performed by a human driver (supervisor) possessing certain specific qualifications, including among others, having possessed a driver’s license for at least five years concerning the specific vehicle under testing, and other requirements related to accreditation in another EU country and training in a third country where such vehicles operate for a specified minimum number of miles. In particular, the regulations require that the human driver be in a condition to promptly switch between the automatic and manual operation of the vehicle, and vice versa. Back to Top The government of Jersey has stated that AI poses both a threat and opportunity to its jurisdiction, particularly with regard to its significant financial sector.
The Dutch government concurs with this view. It therefore proposed, among other things, to establish a GGE at the 2015 annual meeting of the Convention on Certain Conventional Weapons (CCW) to study this issue. The Dutch government emphasizes that it rejects the development of fully autonomous weapon systems without any human control. Back to Top Norway has adopted a national digital agenda, which includes a focus on AI investments. In the agenda the government presents how Norway can best utilize the possibilities of AI and modern technology. In addition Norway signed onto the EU collaboration on AI as part of Digital Day 2018. Norwegian government agencies also use AI in their operations. As noted above, the Nordic countries have also joined together in a number of collaborative responses to AI, including together with the Baltic States. The testing of self-driving cars in Norway is regulated by law. The purpose of the law is to enable the testing of self-driving technology while also factoring in traffic safety and privacy protection. Norway has issued a number of statements on LAWS, but has yet to adopt an outright policy. Speaking more generally on LAWS at the Convention on Certain Conventional Weapons meeting in 2016, the Norwegian representative declared that it “is necessary to ensure that the basic rules and principles of international law are upheld.” Back to Top Poland’s Ministry of Infrastructure and Transport has presented the Operational Program for Digital Poland 2014–2020. The aim of the Program is to strengthen digital foundations for national development, including common “access to a high-speed Internet, effective and user- friendly public e-services and a continuously rising level of digital competences of the society.” The Polish government initiated the development of an AI strategy in 2018 in line with the priorities outlined in the Operational Program for Digital Poland.
On January 11, 2018, the Polish Sejm adopted the Law on Electromobility and Alternative Fuels. Article 65K of the Law defines autonomous vehicle as “a vehicle, equipped with systems that control the movement of the vehicle and enables its motion without driver intervention, which can take control of the vehicle at any time” The Law also stipulates that conducting research of autonomous vehicles for transit of transportation purposes is allowed, provided that the safety conditions are met and mandatory permits are obtained. A designated stand-by driver must be present in the vehicle during the test drive. Back to Top As noted in the European Union survey, the European Parliament and the Council of the European Union approved a new general regulation on the protection of individuals with regard to the processing of personal data and on the free movement of such data on April 27, 2016. The regulation was created to protect citizens from the processing of personal data on a large scale by large companies and information society services. It entered into force in the European Union on May 25, 2018. Portugal has yet to transpose the new regulation into its domestic legal system. Portuguese news sources have noted that the new rules should boost consumer and business confidence. With the new general regulation, it becomes possible to ask a company to reveal all the data it has about a person, erase this data, and review some decisions made by software. Concerns over the protection of digital data and the impact of AI on workers are among the main concerns of several Portuguese companies who would otherwise invest in the area. Portugal is taking the first steps to use AI in the area of justice. A legal assistance tool adapted to the Portuguese reality that conducts research on the requests made and learns from them will soon be launched in the country. Some have speculated that, in the future, with the use of the same technology, it will be possible to launch a tool that predicts the probability of success of a judicial process with the use of AI for the analysis of a repository of judgments issued by the EU with more than three hundred thousand decisions of the Portuguese High Courts. Portugal is using more AI in several projects to analyze data from the public administration in order to find hidden information and create solutions in the areas of health, education, urban mobility, and land use and planning. All initiatives result from cooperation between scientific institutions and entities of the public administration that are being developed under the innovation axis of the National Digital Competencies Initiative, which the government established in 2017 to increase digital skills in the country. Back to Top In 2017, President Putin expressed his opinion that the country that masters AI first will become the “ruler of the world.” Russia’s efforts in the development of AI infrastructure focus on upgrading the country’s capabilities in the area of defense.
These entities are active participants in both national and international efforts to support and promote research and development in AI. As part of the Digital Strategy for a Smart Spain 2025, in November 2017 the Secretary of State for the Information Technology Society and the Digital Agenda created a working group of experts to develop official guidelines on AI. The group includes businesspeople, professors, and government officials who are reputable and well-known experts in their respective areas of knowledge. The findings of the working group will be published in the Libro Blanco de Inteligencia Artificial (White Book on Artificial Intelligence, WBAI), which is aimed at addressing the social, legal, and ethical implications of the use of AI and big data in the private and public sectors as well as the Spanish society in general. The WBAI working group was expected to have a draft completed by May 2018. The WBAI recommendations will be the basis for a code of ethics and a legal framework on the use of AI data in the public administration as well as a related code of good practices for companies, to be prepared by the government. In August 2018, the working group submitted a draft of the WBAI to the new Spanish government headed by Pedro Sanchez. Although the working group, created by the previous Spanish government, has not been dissolved, the deadline for the WBAI publication is not clear, as there has been no official announcement in this regard. In addition, Spain signed the EU Cooperation Agreement on Artificial Intelligence on April 10, 2018. The agreement aims to promote Europe’s technology and industrial capacity in AI, address the challenge that AI poses to the labor market and education and training systems in the EU, and devise the appropriate AI legal and ethical framework, taking into consideration the EU fundamental rights of privacy and the protection of personal data. Although Spain has not enacted any legislation or regulations on autonomous vehicles, the General Directorate of Traffic has authorized the testing of autonomous vehicles since 2015. The government of Spain has declared that its armed forces do not have and are not planning to develop LAWS or weapons based on AI. The Spanish government is not financing any research project on LAWS.
The government further supports the adoption of a general international legal framework aimed at preventing an arms race with respect to LAWS. The Spanish government supports the issuance of the Political Declaration and politically binding Code of Conduct that would include transparency measures to monitor this process as well as the establishment of a Committee of Experts who will be responsible of the evaluation of new developments in the technology involving these weapons. Back to Top Sweden is a member of the European Union and therefore the General Data Protection Regulation (GDPR) on privacy has applied directly since May 2018. >Regulation of AI has also been viewed as a task for the entire EU to address jointly, which is why Sweden has supported a position that the European Commission should consider this. Sweden has also signed on to the Declaration on Cooperation on AI as part of Digital Day in 2018. Despite arguing for an EU approach to AI, the Swedish government and Swedish universities as well as philanthropist are investing in AI. Sweden has developed a National Strategy for Artificial Intelligence. The goal is to become a leader in new digital opportunities. This means creating “the enabling conditions”
Sveriges Kommuner och Landsting (SKL) (the Swedish municipality and county council agency) has issued a report on the potential use of AI in health and well care settings. Other examples of AI use include its use in educational settings and in digital sobriety tests by Kontigocare. SKL also predicts that the use of robots and AI technology could be helpful for the county councils and regions in their goal to provide their residents with välfärd services (collective name for health services, public transportation, education, elder care, child care, water, roads, and other public utilities) in the near future. Studies have also been completed to review the usefulness of AI technology in Swedish industry and enterprise. The Swedish government has issued a government report on self-driving cars in order to “analyze what rule changes are needed for an introduction of driver-supported technology and wholly or partially self-driving vehicles on the road.” One of the issues will be for producers of autonomous vehicles to prove how safe the technology is to justify putting it on the roads. Current legislation allows for the testing of self-driving cars following a permission system. A physical driver must be present inside or outside the car. Permits are awarded by the Transportstyrelsen (Swedish Transport Agency). In addition, Transportstyrelsen has issued its own rules. The rules include the information that must be contained in an application. It also provides that the results of the test must be reported to the Transportstyrelsen when completed. In addition, if tests are run for more than a year, annual reports must be submitted. According to media reports, a total of three companies have been approved to test self-driving technology in Sweden: Nobina Technology AB, Amobility AB, and Volvo Cars. An additional four companies have applications pending. Sweden has not publicly called for a ban on the use of LAWS.
- On 30. oktober 2020
HHS Artificial Intelligence Industry Day Part 1
HHS has launched a large scale "Reimagine HHS-BUYSMARTER Acquisitions" initiative to procure an AI solution that generates efficiencies & provides cost ...
Engaging with China and East Asia
Delaware U.S. Senator Chris Coons and former National Security Advisor Tom Donilon headline a two part discussion on US and China Relations. The event is ...
What Makes a Superpower? #AshCarter #GZW135 #CL0
The United States is the world's undisputed superpower. But how much longer can that last? Enter: China. On the show Ian Bremmer breaks down the three key ...
Consumer Protection in an Age of Uncertainty - Day 2
Consumer Protection in an Age of Uncertainty - Day 2.
Joy Buolamwini on The Open Mind: Algorithmic Justice
On this week's episode of The Open Mind, we welcome Algorithmic Justice League founder Joy Buolamwini. Her organization's mission is to highlight ...
Reconciling International Security
Many effective altruists prioritize reducing existential risks. The international security community also works to lower the odds of certain sorts of existential risks, ...
Revolution OS - 2001 - Multilingual (16 languages)
Subtitles found on OpenSubtitles.org: Arabic, Bulgarian, Chinese, Croatian, Czech, English, French, German, Greek, Polish, Portuguese, Rumanian, Russian, ...
MFA Products of Design 2016: Thesis Show
13:30 Opening Remarks, Allan Chochinov Group 1 36:42 Marianna Mezhibovskaya, Outsiders: Designing Engagement with the Incarcerated 51:24 Adem ...