Page 37 - BIPAR Annual Report 2020_EN
P. 37

in the development of digital responsibility principles in   This Regulation provides for general rules on online   Artificial Intelligence (AI)   accountability. It launched a large-scale pilot phase for
 insurance. The Expert Group aims to address the use of new   platforms, not specifically targeting the financial/insurance   feedback from stakeholders in the summer of 2019. It works
 business models, new BDA tools (such as AI and ML) and   sector, and it does not affect the application of the relevant   Background  on international consensus building for humancentric AI. The
 data sources in insurance from the perspective of fairness   rules of Union law applicable in financial services (e.g. the   HLEG concludes that, while these Guidelines aim to build a
 and non-discrimination,  transparency and  explainability,   IDD). However, these rules on online platforms will apply in   Algorithms are behind more and more decisions that affect our   horizontal framework to achieve Trustworthy AI, it should be
 and governance. Different areas of the insurance value chain   addition to sector-specific legislation when:   everyday lives, for example getting a loan, or the selection of   explored whether a sectorial approach is needed, given the
 will be covered with specific focus to be given to pricing and   filtering of information. In several areas, there are already EU   context-specificity of AI systems, and that the Guidelines are
 underwriting. The two candidates that BIPAR had supported   -   Insurance/financial  intermediaries  are  business   rules for algorithmic decisions. Examples include automated   a living document which should be reviewed and updated
 were  both  appointed  members  of  the  Expert  Group.  The   users of online platforms/corporate websites in order to   decisions  based  on  personal  data  (GDPR)  and  for  high-  over time to ensure continuous relevance.
 Group is expected to adopt a report and some proposed   offer (ancillary) insurance/financial services; in other words,   frequency trading on the stock market (MiFID II).
 principles in the development of digital responsibility in   they will be entitled to the rights provided in the Regulation.  EU Digital Strategy 2020
 insurance by the end of 2020.  -   Insurance/financial   intermediaries   (including   In  April  2018,  the  European  Commission  proposed  in  its   On 19 February 2020 the European Commission announced
 comparison  tools)  are  providers  of  online  platforms/  “Strategy  on  AI  for  Europe”  to  work  with  Member  States   the  initiatives  and  actions  to  be  taken  over  the  next  five
 BIPAR believes that intermediaries, who are daily confronted   corporate  website  offering  (ancillary)  insurance/financial   on  a  Coordinated  Plan  on  AI  with  the  aim  of  increasing   years in order to address the challenges and opportunities
 with  the  processing  of  consumer’s  data,  should  actively   services; in other words, they should comply with the   investments in AI research and innovation,   brought about by digitalisation. The
 explore the potential of the use of BDA in order to improve   transparency obligations laid down in the Regulation.  encouraging  synergies  and  cooperation   EU Digital Strategy is presented in the
 their processes. BIPAR also emphasises the need for a level   across  the  EU,  preparing  socio-economic   Commission Communication  “Shaping
 playing  field  among  the  different  distribution  channels   changes brought about AI and ensuring an   Europe’s Digital Future”.
 and warns of the challenges arising with regard to the   appropriate ethical and legal framework.
 mutualisation principle, the comparability of products and                   The   Communication  presents  the
 the cyber risks. The development and use of cyber insurance   According   to   the   Commission’s   three  key  objectives  on  which  the
 products may be one of the answers for the prevention of   Communication   on   AI,   “artificial   Commission will focus to promote digital
 such risks.  intelligence refers to systems that display                     transformation for the benefit of people:
            intelligence  behaviour  by  analysing  their                     •            Technology that  works for
 Transparency of online intermediation services  environment  and  taking  actions  –  with   people;
            some  degree  of  autonomy  –  to  achieve                        •           A fair and competitive
 Background  specific goals”. The High-Level Expert Group (HLEG) on AI set   economy; and
            up by the Commission published in December 2018 a paper   •        An open, democratic and sustainable society
 In April 2018, the European Commission presented a proposal   The  Regulation  introduces  a  ban  on  certain  unfair   determining the definition of AI in more detail.
 for  a  Regulation  of  the  European  Parliament  and  of  the   practices (e.g. terms and conditions must be easily
 Council on promoting fairness and transparency for business   available and provided in plain and intelligible language),   The HLEG on AI published its “Ethics Guidelines for   The White Paper on Artificial Intelligence and the European
 users of online intermediation services.  greater  transparency  in  online  platforms  (e.g.  main   Trustworthy AI” in April 2019.  According to the Guidelines,   Data Strategy presented together with the EU Digital
 parameters determining ranking, including direct or   trustworthy AI has three components, which should be met   Strategy  are  the  first  steps  towards  achieving  these  goals.
 The Regulation on promoting fairness and transparency for   indirect remuneration paid by business users), mandatory   throughout  the  system’s  entire  life  cycle:  1)  it  should  be   The White Paper on AI sets out the Commission’s proposals
 business users of online intermediation services has been   disclosure for a range of business practices (platforms   lawful, complying with all applicable laws and regulations 2)   for a trustworthy and secure development of AI in Europe.
 published in the Official Journal of the EU in 20 June 0219   must exhaustively disclose any advantage they may give to   it should be ethical, ensuring adherence to ethical principles   The Commission presents policy options on creating a
 and it will start to apply in EU Member States from 12 July   their own products over others), new avenues for dispute   and values and 3) it should be robust, both from a technical   legal framework that addresses the risks for fundamental
 2020 (Regulation 2019/1150). All language versions of this   resolution (all platforms must set up an internal complaint-  and social perspective since, even with good intentions, AI   rights and safety, including in the field of financial services.
 Regulation can be found here. The Regulation aims to ensure   handling system to assist business users) enforcement rules   systems can cause unintentional harm.  The  Commission  Communication  on  the  “European  Data
 that  business  users  of  online  intermediation  services  and   (business associations will be able to take platforms to court   Strategy”  outlines  a  strategy  for  policy  measures  and
 online search engines are granted appropriate transparency,   to stop any non-compliance with the rules).  The  Guidelines  offer  guidance  on  the  second  and  third   investments to enable the data economy for the coming five
 fairness and effective redress possibilities. The Regulation   To assist online intermediation services and online search   component of trustworthy AI. The Commission is taking a   years. The ultimate objective is to set up a true European
 covers online platforms, such as price comparison websites,   engines in complying with the obligation to set out in their   three-step approach. It sets out the key seven requirements   data space, a single market for data (personal and non-
 and online search engines that provide their services to   terms and conditions the main parameters determining   that AI systems should meet: Human agency and oversight,   personal  data,  including  confidential  and  sensitive  data),
 businesses which are established in the EU and which offer   ranking and the reasons for the relative importance of those   Technical  robustness  and  safety,  Privacy  and  data   to  unlock  unused  data,  allowing  it  to  flow  freely  within
 goods or services to consumers located in the EU.  main parameters, the European Commission is planning to   governance, Transparency, Diversity, non-discrimination   the European Union and across sectors. The Commission
 adopt guidelines.  and  fairness,  Societal  and  environmental  well-being  and   proposes, amongst other, to Launch sectoral specific actions



 36                                                         37
   32   33   34   35   36   37   38   39   40   41   42