Posted: March 19 2020 on Unissu with the following editor’s Note:
This is one of the most complex areas of AI, and certainly AI in real estate: How do we regulate its use, and who is responsible for doing so? Complex not only because regulation needs to constantly match invention, but because the strength and ability of AI in real estate is going to keep moving forward at an amazing rate. Milan Bogár takes on the challenge of trying to figure out who this responsibility should fall to.
It’s awesome where we are in the world of tech… Scary, too. These opposite words from ‘profound’ pop musician will.i.am resounded, said during one of the latest parts of the YouTube series – the Age of A.I. 1.st (How far is too far?).
Will.i.am announces these words before he dares to speak to his own AI twin, created by Soul machines, one of the most advanced companies developing purely digital workforce. Do you want to hire purely digital workforce? No problem. It is not science fiction any more.
Watching the last videos with Sofia, the first AI robot to receive citizenship of a country is, again, as fascinating as it is worrying. I love her sense of humour. Would you believe, that she became a citizen of Saudi Arabia in 2017?
AI as subject of civil rights become reality almost three years ago. Since October 2017, the United Arab Emirates are the first country in the world to have the Ministry of State for Artificial Intelligence. Can we assume that this is the watcher over the watchers that we are looking for?
United Arab Emirates are the first country in the world to have the Ministry of State for Artificial Intelligence
AI inventions can be subject of protected intellectual property. This year, it is 64 years since the term Artificial intelligence was used for the first time. It is unbelievable that, since that time, innovators and researchers have filed patent applications for more than 340,000 AI-related inventions worldwide according to the WIPO.
The increase in computer power, amount of data, and connectedness has brought an enormous boom and new breakthroughs especially in the last decade. More than half of all AI patents in history was filed after 2012. Over the period 2013-16, the most popular functional applications have been those encompassing computer vision (49%) which is critical for instance for self-driving cars or entry and access systems based on face recognition, then natural language processing (14%) and speech processing (13%). The vast majority of these have been attributed to human inventors.
But, inventions can already be autonomously generated by AI and there are several reported cases of applications for patent protection in which the applicant has named an AI application as the inventor!
Then, who is responsible for AI regulating use of AI in real estate specifically? We have various kinds of responsible regulatory authorities across all sectors where AI can be used. There are different regimes for protection of fundamental rights, for intellectual property rights, consumer protection rights, personal data, cybersecurity and different sector specific regulatory regimes related to healthcare, energy, transportation, electronic communications and others.
We have various kinds of responsible regulatory authorities across all sectors where AI can be used…but I did not find one that would tackle the regulation of AI in RE
But, I did not find one that would tackle the regulation of AI in real estate sector specifically. I think it is because real estate is one of the most interdisciplinary sectors and the same authorities are responsible for regulation of AI in real estate as in each of the sectors for which the real estate actually serves.
Authorities responsible for regulating protection of personal data are enforcing compliance with respective personal data obligations related to AI across all sectors. For lawful usage of AI application for processing personal data, it is extremely important the following obligation is adhered to: “Controllers must, at the time when the personal data are obtained, provide the data subjects with further information necessary to ensure fair and transparent processing about the existence of automated decision-making and certain additional information” (Art. 13(2)(f) GDPR).
If the controller is the real estate owner, they must provide the transparent information to all data subjects whenever their data are obtained for processing with assistance of AI. But is it possible in all AI applications?
On 19 February 2020, European Commission has issued long-awaited White Paper on Artificial intelligence – A European approach to excellence and trust.
Here is just a small taster of the very interesting document:
Why an approach to excellence and trust? Excellence because European executives clearly assume that the responsible development and use of AI can be a driving force in achieving the Sustainable Development Goals. But they are also aware of the lack of private and public investment and lack of skills in that need to be overcome in order to unleash the excellent force of AI in European union. And trust, because lack of trust is a main factor holding back a broader uptake of AI.
Lack of trust is a main factor holding back a broader uptake of AI.
The Commission believes that “A solid European regulatory framework for trustworthy AI will protect all European citizens and help create a frictionless internal market for the further development and uptake of AI as well as strengthening Europe’s industrial basis in AI.
In the view of the European Commission, the risk-based approach is important for ensuring that the regulatory interventions related to AI are proportionate. In this approach only those AI applications that imply ‘high-risk’ should need new and more strict regulatory requirements encompassing for instance additional testing and certification. Those AI applications that do not imply high-risk should remain entirely subject to already existing EU-rules with possibility of new voluntary labelling. The European Commission proposes easily understandable differentiation criteria as well.
The new requirements for high-risk AI applications could consist of the following key features:
(1) training data (data used for learning AI, need to be reliable, complete, not to result in garbage in – garbage out or otherwise biased or discriminatory decision-making process);
(2) data and record-keeping (to address opacity in potentially problematic actions or decisions by AI systems the underlying processes has to be enable tracing back and transparent verification);
(3) information to be provided (citizens should be clearly informed when they are interacting with an AI system and not a human being and about capabilities and limitations of the AI);
(4) robustness and accuracy (as minimum requirements to come to accurate, reproducible results and dealing with errors or inconsistencies during all life cycle phases);
(5) human oversight (enabling human monitoring, intervention, validation or approval before or after use, or in the design phase of AI applications);
(6) specific requirements for certain particular AI applications, such as those used for purposes of remote biometric identification (possible only if duly justified, proportionate and subject to adequate safeguards).
Finally, the White Paper proposes a European governance structure on AI in the form of a framework for cooperation of competent authorities in EU and member states. This governance structure should avoid fragmentation of responsibilities, increase capacity in Member States, and make sure that Europe equips itself progressively with the capacity needed for testing and certification of AI-enabled products and services.
Despite the lack of real estate perspective in the White Paper, AI applications are clearly becoming one of the most important property technologies in the next decade. This reminds me the objectives of the PropTech House – The Alliance of European PropTech Associations:
“Our objectives: to scale up cross-border collaborations, to standardise the European PropTech markets, to create a legal framework adapted to PropTech, to foster innovation in Real Estate and to facilitate access to funding.”
Specifically, new funding is proposed in Action No. 4 of the White Paper: The Commission and the European Investment Fund will launch a pilot scheme of €100 million in Q1 2020 to provide equity financing for innovative developments in AI. Subject to final agreement on the MFF, the Commission’s intention is to scale it up significantly from 2021 through Invest EU.
The White Paper consultation is open for comments until 19 May 2020 and all PropTech associations and players have good chance to participate as well as all the other stakeholders.