{"id":7181,"date":"2023-12-18T09:04:56","date_gmt":"2023-12-18T09:04:56","guid":{"rendered":"https:\/\/members.dataiq.global\/?post_type=article&#038;p=7181"},"modified":"2024-03-19T09:00:32","modified_gmt":"2024-03-19T09:00:32","slug":"developing-and-introducing-framework","status":"publish","type":"article","link":"https:\/\/www.dataiq.global\/devstage\/articles\/developing-and-introducing-framework\/","title":{"rendered":"Developing and introducing an ethical AI framework"},"content":{"rendered":"<p><span data-contrast=\"auto\">Ethical frameworks for AI follow a similar approach\u00a0and the lessons learnt from GDPR implementation can be used as a guide for this new era of data.<\/span><\/p>\n<p>&nbsp;<\/p>\n<h4><strong style=\"font-weight: 600;\"><span data-contrast=\"auto\">Starting the framework\u00a0<\/span><span data-ccp-props=\"{\">\u00a0<\/span><\/strong><\/h4>\n<p><span data-contrast=\"auto\">The\u00a0Organisation for Economic Co-operation and Development\u00a0(OECD)\u00a0produced its<em>\u00a0<\/em><\/span><a style=\"font-weight: 600;\" href=\"https:\/\/oecd.ai\/en\/ai-principles\" rel=\"nofollow noopener\" target=\"_blank\"><span data-contrast=\"none\"><em>Principles for Ethical AI<\/em>,<\/span><\/a>\u00a0<span data-contrast=\"auto\">which were adopted in May 2019.\u00a0Alongside the OECD\u2019s\u00a0series of values-based principles,\u00a0were a published list of\u00a0recommendations for policy\u00a0makers:<\/span><span data-ccp-props=\"{\">\u00a0<\/span><\/p>\n<p><em><span data-contrast=\"auto\">Values-based principles<\/span><\/em><\/p>\n<ul>\n<li><span data-contrast=\"auto\">Inclusive growth, sustainable development and well-being<\/span><\/li>\n<li>Human-centred values and fairness<\/li>\n<li>Transparency and\u00a0explainability<\/li>\n<li>Robustness,\u00a0security\u00a0and safety<\/li>\n<li>Accountability<\/li>\n<\/ul>\n<p>&nbsp;<\/p>\n<p><em>Recommendations for policy makers<\/em><\/p>\n<ul>\n<li><span data-contrast=\"auto\">Investing in AI\u00a0research and development<\/span><\/li>\n<li>Fostering a digital ecosystem for AI<\/li>\n<li>Providing\u00a0an engaging policy environment for AI<\/li>\n<li>Building human\u00a0capacity\u00a0and preparing for labour market transition<\/li>\n<li>International cooperation for trustworthy AI<\/li>\n<\/ul>\n<p>&nbsp;<\/p>\n<p>In April 2021, the European Commission proposed the first EU regulatory framework for AI,\u00a0stating\u00a0that AI systems used in different applications should be analysed and classified according to the risk they pose to users.\u00a0The different risk levels will\u00a0indicate\u00a0the amount of regulation\u00a0required.\u00a0<span data-ccp-props=\"{\">\u00a0<\/span><\/p>\n<p><span data-ccp-props=\"{\">\u00a0<\/span><\/p>\n<h4><strong style=\"font-weight: 600;\"><span data-contrast=\"auto\">Implementing framework internally <\/span><span data-ccp-props=\"{\">\u00a0<\/span><\/strong><\/h4>\n<p><span data-contrast=\"auto\">Data leaders need to\u00a0locate\u00a0pockets of the enterprise where teams are considering use cases of AI and\u00a0reach out for early adopters of any framework being suggested. By building\u00a0the importance of an ethical approach, a stronger level of success and implementation will be achieved and set the path for continued data culture evolution embracing an ethical AI framework.\u00a0<\/span><span data-ccp-props=\"{\">\u00a0<\/span><\/p>\n<p><span data-contrast=\"auto\">Organisations using AI should develop an AI risk impact assessment framework, which\u00a0should have\u00a0similarities to data protection risk impact assessments. Software tools\u00a0can\u00a0support this process which typically involves\u00a0answering questions relating to the use case, with high-,\u00a0medium-\u00a0and\u00a0low-risk\u00a0consequence answers.<\/span><span data-ccp-props=\"{\">\u00a0<\/span><\/p>\n<p><span data-contrast=\"auto\">Data leaders must\u00a0consider the degree to which the use case has autonomous decision making once the human is no longer in the loop.\u00a0Naturally, if there\u00a0would be\u00a0legal\u00a0impacts\u00a0of the decision, the risks rise considerably.\u00a0Additionally, if there\u00a0would\u00a0be considerable reputational damage\u00a0it is important to\u00a0identify\u00a0the best ways to mitigate the risk.\u00a0<\/span><span data-ccp-props=\"{\">\u00a0<\/span><\/p>\n<p><span data-contrast=\"auto\">It is\u00a0important to keep revisiting the risks once they have been\u00a0identified\u00a0\u2013 it is\u00a0not a one-time situation as risks evolve. Data leaders should complete\u00a0several risk impact assessments when the team are building the product and another when the ultimate use is defined.<\/span><span data-ccp-props=\"{\">\u00a0<\/span><\/p>\n<p><span data-contrast=\"auto\">Many companies are introducing AI\u00a0forums which have representatives from across the business, which is reflective of the data protection officer\u00a0committees\u00a0that\u00a0emerged\u00a0in 2016. Procurement and IT are critical functions\u00a0to be addressed from an ethical framework point of view\u00a0as most of these applications require either internal technical resources or external vendor support.\u00a0Data leaders should consider asking all vendors for their\u00a0own\u00a0ethical AI policies\u00a0which\u00a0can\u00a0be included in vendor contracts.<\/span><span data-ccp-props=\"{\">\u00a0<\/span><span data-ccp-props=\"{\">\u00a0<\/span><\/p>\n<p><span data-contrast=\"auto\">For teams to\u00a0comprehensively\u00a0understand how to implement ethical AI, it is\u00a0important for there to be a degree of\u00a0training, which has\u00a0similarities to the\u00a0initial\u00a0implementation of\u00a0GDPR training.\u00a0<\/span><span data-ccp-props=\"{\">\u00a0<\/span><\/p>\n<p><span data-contrast=\"auto\">Ethical frameworks are nothing new. However,\u00a0as more\u00a0companies experiment\u00a0with AI\u00a0\u2013\u00a0which removes the human\u00a0\u2013\u00a0the need to consider consequences has\u00a0forced a review of the approach to ethical data processing. Those with strong ethics as part of their values have found this a key benefit in making\u00a0AI\u00a0implementation\u00a0decisions.<\/span><span data-ccp-props=\"{\">\u00a0<\/span><\/p>\n","protected":false},"excerpt":{"rendered":"<p>Data leaders looking to introduce an ethical AI framework can look back to the way GDPR was implemented to provide guidance on how it can be approached.<\/p>\n","protected":false},"author":19,"featured_media":7182,"menu_order":0,"comment_status":"open","ping_status":"closed","template":"","format":"standard","meta":{"_acf_changed":false,"content-type":"","_searchwp_excluded":"","footnotes":""},"categories":[129,131],"tags":[217,218,237,316,315,317,301,318,304],"pillar":[198,193,194],"class_list":["post-7181","article","type-article","status-publish","format-standard","has-post-thumbnail","hentry","category-editorial","category-member-articles","tag-ai","tag-artificial-intelligence","tag-data-leaders","tag-framework","tag-gdpr","tag-oecd","tag-regulation","tag-training","tag-upskilling","pillar-governance","pillar-strategy","pillar-leadership"],"acf":[],"publishpress_future_action":{"enabled":false,"date":"2026-05-20 23:39:43","action":"change-status","newStatus":"draft","terms":[],"taxonomy":"category","extraData":[]},"publishpress_future_workflow_manual_trigger":{"enabledWorkflows":[]},"_links":{"self":[{"href":"https:\/\/www.dataiq.global\/devstage\/wp-json\/wp\/v2\/article\/7181","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.dataiq.global\/devstage\/wp-json\/wp\/v2\/article"}],"about":[{"href":"https:\/\/www.dataiq.global\/devstage\/wp-json\/wp\/v2\/types\/article"}],"author":[{"embeddable":true,"href":"https:\/\/www.dataiq.global\/devstage\/wp-json\/wp\/v2\/users\/19"}],"replies":[{"embeddable":true,"href":"https:\/\/www.dataiq.global\/devstage\/wp-json\/wp\/v2\/comments?post=7181"}],"version-history":[{"count":0,"href":"https:\/\/www.dataiq.global\/devstage\/wp-json\/wp\/v2\/article\/7181\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.dataiq.global\/devstage\/wp-json\/wp\/v2\/media\/7182"}],"wp:attachment":[{"href":"https:\/\/www.dataiq.global\/devstage\/wp-json\/wp\/v2\/media?parent=7181"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.dataiq.global\/devstage\/wp-json\/wp\/v2\/categories?post=7181"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.dataiq.global\/devstage\/wp-json\/wp\/v2\/tags?post=7181"},{"taxonomy":"pillar","embeddable":true,"href":"https:\/\/www.dataiq.global\/devstage\/wp-json\/wp\/v2\/pillar?post=7181"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}