点击上方蓝字关注我!



编者按:
2020年2月19日,欧洲委员会发布了《人工智能白皮书》《数据战略报告》和《数字未来报告》。本期知竞团队翻译了比利时律师Kyriakos Fountoukakos于2020年3月6日在Mondaq网站上就三份报告发布的评论文章,以保持对人工智能以及数字市场规制问题的持续关注。文章原文请点击“阅读原文”获取。
European Commission Announces Strategy For Data, Artificial Intelligence And Competition In The Digital Age
欧盟委员会发布数字时代下的数据、人工智能和竞争战略
2020年2月19日,欧洲委员会发布了三份作为其“塑造欧洲数字未来”战略的一部分的政策文件: 一份关于人工智能的白皮书(《人工智能白皮书》)、一份关于欧洲数据战略的报告(《数据战略报告》)和一份关于塑造欧洲数字未来的报告(《数字未来报告》)。
这些文件中所载的建议和倡议是委员会在“适合数字时代的欧洲”项目中作出的第一项重要公告,这是冯 · 德莱恩(von der Leyen)主席在2019-2024年政治指导方针中提出的六项主要目标之一。该公告也是玛格丽特·韦斯特格(Margrethe Vestager)担任欧盟委员会执行副主席与负责“适合数字时代的欧洲”项目和竞争事务专员的双重角色以来,该委员会在该领域的第一项重大监管提案。从竞争法的角度来看,《数据战略报告》和《数字未来报告》载有以下方面的建议:对“大型高科技”平台进行事先管制;可能更新竞争法以适用于数字市场;如何在兼并控制分析中考虑到数据的收集和使用;以及自愿和(酌情)强制性数据共享。这些文件似乎是雄心勃勃的,其中有多少建议和想法最终将在中短期内得到落实还有待观察。
人工智能白皮书
《人工智能白皮书》阐述欧洲委员会建议的人工智能监管方法。重点是采用以风险为本的方法,对“高风险”应用程序施加强制性监管要求,即将人工智能部署在可以预见重大风险的行业(例如医疗保健)中,并且在可能导致重大风险的行业应用人工智能。对于高风险(和某些其他)应用程序,《人工智能白皮书》提出了以下方面的要求:培训数据;数据和记录保存;提供的信息;稳健性和准确性;人为监督;以及对某些特定人工智能应用的具体要求,例如用于远程生物识别的要求。至于非高风险的人工智能应用,《人工智能白皮书》建议推行自愿性标签计划,通过该计划,申请开发人工智能应用的经营者将被授予“质量标签”。
《人工智能白皮书》还提出了有关新监管框架的合规性、强制性和治理方面的建议,尤其包括使用先前的合格性评估(在欧盟内部市场投放的许多产品已经采用),以及适用于人工智能应用程序的测试、检查或认证程序,包括潜在地检查在开发阶段使用的算法和数据集。最后,《人工智能白皮书》提出可能调整欧盟产品安全和责任立法,以及可能专门针对人工智能立法,以便使欧盟的法律框架适应当前和预期的技术和商业发展(《人工智能白皮书》附带的另一份报告进一步探讨了这一点)。
人工智能白皮书将开放征询意见直至2020年5月19日。
数据战略报告
《数据战略报告》提出了一系列旨在促进欧盟数字经济的举措。除此之外,《数据战略报告》还提出了一些可能的立法倡议(作为《数据法案》的一部分) 以促进企业和企业之间的数据共享。作为这项工作的一部分,委员会将就数据共享和汇总是否符合欧盟竞争法提供指导。委员会还将考虑在“公平、透明、合理、相称和/或非歧视的条件下”在特定情况下强制共享数据。
《数据战略报告》讨论了“大型高科技”公司积累的大量数据,这可能会产生竞争优势,从而影响市场的竞争能力,并使这些公司的优势地位相对容易地传播到周边市场。但是,此问题不属于《数据法案》的一部分,而是将在《数字未来报告》”的提案或倡议下进行审核(请参见下文)。
关于合并控制,《数据战略报告》指出,委员会将研究通过收购积累大规模数据对竞争可能造成的影响,以及利用数据访问或数据共享补救措施来解决问题。关于国家援助,数据战略报告指出,委员会将在对一些国家援助准则的持续审查中,研究“公共支持对企业(例如,数字化转型企业)与通过以下方式将竞争最小化的关系:受益人的数据共享要求。”
其他建议包括:
根据经修订的《公共部门信息再利用的开放数据指令》界定公共部门“高价值数据集”的实施方法,以便对此适用更大的访问权;
解决与共同生成的数据(例如工业环境中的物联网数据)使用权有关的问题并提出规制用于数据分析和机器学习的数据池的可能的立法倡议;以及
在公共利益的战略部门和领域中建立“欧洲共同数据空间”,目的是在这些部门中创建大量可用数据,并结合使用和交换数据所必需的技术工具和基础架构以及适当的治理机制。列出的部门包括:工业、“欧洲绿色交易”区域、交通、健康、金融、能源、农业、公共管理。
委员会目前正在收集有关其数据战略的反馈意见(似乎尚未确定截止日期)。
数字未来报告
《数字未来报告》提出了三个关键原则和目标:(1)“为人民服务的技术”,(2)“公平竞争的经济”和(3)“开放、民主和可持续的社会”。文中包含一系列促进这些原则和目标的建议,但是在很大程度上没有涉及具体建议的细节。关于第二个目标,委员会强调必须确保欧盟单一市场中企业的公平竞争环境。作为这项倡议的一部分,委员会将:
于2020年对数字市场进行行业调查;
持续评估(或审查)欧盟竞争规则对数字时代的适用性(直至2023年)。委员会已经在审查关于横向和纵向协议以及市场定义的规则;
探索引入事前监管的可能性,以确保“以大型平台为代表的网络具有巨大网络效应的市场,对创新者、企业和新进入市场者保持公平和可竞争”。和
致力解决经济数字化所带来的税务挑战。
其他两个目标与竞争政策的关系较小。与第一个目标相关的建议包括促进对欧洲千兆连接的投资、为互联互通部署5G走廊、提高数字素养和技能以及改善平台工作者的劳动条件,而与第三个目标相关的建议包括协调统一在线平台在其托管内容方面的责任、修订有关电子交易电子识别和信任服务的eIDAS法规等。
评论
欧盟委员会的主要目标之一(特别是与数据战略和数字未来相关的方面)是在数字时代下访问消费者和数据方面为欧洲中小型企业维持公平竞争的环境。从这个意义上说,委员会似乎建议采取双管齐下的办法,以促进形成公平竞争环境。
首先,强制性的数据共享义务可以设法确保小型企业能够获得与大型企业相同的数据(与委员会正在对亚马逊进行的调查相同,委员会指控亚马逊滥用其市场地位,收集商家的销售数据,并利用这些数据为自己谋利)。其次,我们还可以看到关于通过数字平台分销其产品或服务的企业的数字平台处理的事先规定。这些提议尤其针对的似乎是“大型高科技”平台,它通常指Google、Apple、Facebook、Amazon,有时还包括Microsoft(不排除其他的应用程序)。 这种监管干预措施可以取代适用于这一领域的现行竞争法制度,避免进行旷日持久和复杂的反垄断调查(例如谷歌购物案和谷歌安卓案已经完成的上诉调查,以及正在进行的对亚马逊数据做法的调查)。
然而,最重要的是,这些报告为未来数字领域(尤其是与人工智能和数据相关)的监管计划提供了指引。这些文件似乎雄心勃勃,最终在中短期内提出多少建议和想法尚待观察。有关建议的更多细节在2020年晚些时候发布。
On 19 February 2020, the European Commission (the "Commission") published three policy papers as part of its strategy to "shape Europe's digital future": a white paper on artificial intelligence (AI) (the "AI White Paper"), a communication on a European strategy for data (the "Data Strategy Communication"), and a communication on shaping Europe's digital future (the "Digital Future Communication").
The proposals and initiatives contained in these papers represent the first major announcement by the new Commission in relation to "a Europe fit for a digital age", one of the six primary ambitions set out in President von der Leyen's political guidelines for 2019-2024. The announcement also represents the Commission's first major regulatory proposal in this area since Margrethe Vestager commenced her dual-role as Executive Vice-President of the Commission responsible for "a Europe fit for a digital age" and Commissioner for Competition. From a competition law perspective, the Data Strategy and Digital Future Communications contain proposals relating to: ex ante regulation of "Big Tech" platforms; potentially updating competition law as it applies to digital markets; how the collection and use of data can be factored into in merger control analyses; and voluntary and (where appropriate) compulsory data sharing. The scope of the papers seems to be ambitious, and it remains to be seen how many of the proposals and ideas will ultimately be taken forward in the short to medium term.
AI White Paper
The AI White Paper sets out the Commission's proposed approach to AI regulation. The focus is on a risk-based approach, imposing mandatory regulatory requirements for "high-risk" applications, i.e. where the AI is deployed in a sector where significant risks can be expected (e.g. healthcare) and where risks are likely to arise given the manner of use of the AI (e.g. health risks). For high risk (and certain other) applications, the AI White Paper proposes requirements in relation to: training data; data and record-keeping; information to be provided; robustness and accuracy; human oversight; and specific requirements for certain particular AI applications, such as those used for purposes of remote biometric identification. For non-high risk AI applications, the AI White Paper proposes a voluntary labelling scheme, through which economic operators that opt-in would be awarded a "quality label" for their AI applications.
The AI White Paper also sets out proposals in relation to compliance, enforcement and governance with respect to the new regulatory framework, including in particular the use of prior conformity assessments (which is already in place for many products placed on the EU internal market), with procedures for testing, inspection or certification as adapted to AI applications, including potentially checks of the algorithms and of the data sets used in development phases. Finally, the AI White Paper raises possible adjustments to EU product safety and liability legislation and potentially new legislation specifically for AI, in order to make the EU legal framework fit for current and anticipated technological and commercial developments (this is further explored in a separate report accompanying the AI White Paper).
The AI White Paper will be open to consultation until 19 May 2020 (the consultation will be made available here).
Data Strategy Communication
The Data Strategy Communication sets out a number of initiatives that are essentially intended to promote the EU's data economy. Amongst other things, the Data Strategy Communication proposes possible legislative initiatives (as part of a "Data Act") to promote business-to-business data sharing on a voluntary basis. As part of this, the Commission will provide guidance with respect to the compliance of data sharing and pooling arrangements with EU competition law. The Commission will also consider making data-sharing compulsory in specific circumstances under "fair, transparent, reasonable, proportionate and/or non-discriminatory conditions."
The Data Strategy Communication discusses "Big Tech" companies accumulating large amounts of data, which can result in competitive advantages that affect the contestability of markets and enable those firms to spread into neighbouring markets relatively easily. However, this issue does not form part of the "Data Act" – instead it will be examined under the proposals/initiatives in the Digital Future Communication (see below).
In relation to merger control, the Data Strategy Communication states that the Commission will look at the possible effects on competition of large-scale data accumulation through acquisitions and at the utility of data-access or data-sharing remedies to resolve any concerns. With respect to state aid, the Data Strategy Communication states that the Commission will examine, in its ongoing review of a number of state aid guidelines, "the relationship between public support to undertakings (e.g. for digital transformation) and the minimisation of competition distortions through data-sharing requirements for beneficiaries."
Other proposals include:
an implementing act defining the public sector "high-value data sets" under the revised Open Data Directive on public sector information re-use, for which greater access rights apply;
possible legislative initiatives addressing issues related to usage rights for co-generated data (such as internet of things data in industrial settings), and establishing data pools for data analysis and machine learning; and
the establishment of "common European data spaces" in strategic sectors and domains of public interest, with the aim of creating large pools of available data in these sectors combined with the technical tools and infrastructures necessary to use and exchange data, as well as appropriate governance mechanisms. The sectors listed are: industrial, the "European green deal" area, mobility, health, financial, energy, agriculture, public administration and skills.
The Commission is currently gathering feedback on its data strategy (no deadline appears to have been set).
Digital Future Communication
The Digital Future Communication sets out three key principles/objectives: (1) "technology that works for people", (2) "a fair and competitive economy", and (3) "an open, democratic and sustainable society". The communication contains an array of proposals to promote these principles/objectives, but is largely light on detail in relation to specific proposals. In relation to the second objective, the Commission emphasises the need to ensure a level playing field for businesses in the EU single market. As part of this initiative, the Commission will:
launch a sector inquiry into digital markets in 2020;
evaluate/review the fitness of EU competition rules for the digital age on an ongoing basis (provisionally until 2023). The Commission is already reviewing rules governing horizontal and vertical agreements and market definition;
explore the possibility of introducing ex ante regulation to ensure that "markets characterised by large platforms with significant network effects acting as gate-keepers, remain fair and contestable for innovators, businesses, and new market entrants."; and
look to address the tax challenges arising from digitisation of the economy.
The other two objectives are less related to competition policy. Examples of proposals relating to the first objective include promoting investments in Europe's Gigabit connectivity, rolling out 5G corridors for connected and automated mobility, boosting digital literacy and skills, and improving the labour conditions of platform workers, while examples relating to the third objective include harmonising the responsibilities of online platforms in respect of the content they host, revising the eIDAS Regulation on electronic identification and trust services for electronic transactions, amongst other things.
Comment
One of the Commission's primary aims (in particular in relation to the Data Strategy and Digital Future Communications) is maintaining a level playing field for small-to-medium sized European businesses in terms of access to both consumers and data in the digital age. In this sense, the Commission appears to be suggesting a two-pronged attack to promote a level-playing field.
Firstly, mandatory data-sharing obligations could seek to ensure that small businesses have access to the same data as larger incumbents (in the same vein as the Commission's ongoing investigation into Amazon, where the Commission alleges that Amazon abuses its market position by collecting merchants' sales data and exploiting it to its own advantage). Secondly, we could also see ex ante regulation regarding the treatment by digital platforms of businesses that distribute their products/services through them. These proposals in particular seem to be aimed at "Big Tech", which is generally taken to mean Google, Apple, Facebook, Amazon and sometimes also Microsoft (although a wider application cannot be ruled out). Such regulatory intervention could supplant the current competition law regime as it applies in this area, avoiding the need for drawn-out and complex antitrust investigations (such as the concluded investigations – pending on appeal – in Google Shopping and Google Android and the ongoing investigation into Amazon's data practices).
More than anything else, however, the papers act as a roadmap for future regulatory initiatives in the digital sector, and in particular in relation to AI and data. The scope of the papers seems to be ambitious, and it remains to be seen how many of the proposals and ideas will ultimately be taken forward in the short to medium term. More detail on the proposals that are taken forward should become available later in 2020.




涉及文章内容转载等事宜请联系
微信公众号后台编辑:刘梦佳
审核人:赵丰
邮箱:cptlaw@163.com
手机:15357562155

