In a recent article Dr Malek Rashid referenced the ancient Arabic term “Firasah”: a perceptiveness to see deep into the future. He postulates that, in secular terms, the use of mathematics and algorithms which give meaning to data has the same outcome. This ‘predictiveness’ is a principal feature of Artificial Intelligence (AI) as we are starting to understand it today.
This collation and processing of personal data impacts as never before upon rights to privacy and presents a challenge to Islamic thought and teaching as more and more personal data becomes part of the AI matrix.
Privacy is an essential aspect of life, in both the physical and online worlds. Whether it is traced to Madison’s 1789 Bill of Rights: “Right of the people to be secure in their persons, houses, papers and effects, against unreasonable searches and seizures.” Or whether it is the world’s most powerful companies, using Artificial Technology (AI) to harvest personal information; privacy remains a key social issue.
According to a recent PWC report, AI will impact the Middle East to the amount of US$320 billion in the next 10 years. Saudi Arabia, for example, has now appointed a State Minister for AI as it projects that its economy will benefit to the tune of some US$132 billion over the next several years.
We are all now ‘living online’ through digital platforms and as such new privacy concerns have arisen, in most part from the impact of AI on digital platforms.
With digital platforms implementing AI models such as deep learning, billions of citizens personal information can now be structured, quantified and represented in a form which data analysts can derive insights from: eg Facebook had 100 million active profiles in 2008, now it has an estimated 2.38 billion.. This means a quarter of the planet’s personal information is routinely collected, stored, shared with third parties .
As the Middle East calibrates its approaches to on line AI privacy the definition of what constitutes ‘personal information’ and how we protect it present real challenges : how is modern ‘privacy’ to be reconciled with the rights of privacy evident in the verses of the Holy Quran :
“do not spy on one another.” (49:12)
“do not enter any houses except your own homes unless you are sure of their occupants consent.” (24:27)
An individual’s privacy about their possessions, for example, may easily lead to the identity of the individual with the use of AI : if it were known that only one individual owned a certain brand of mobile device in a class, a simple digital platform survey would lead to their identity.
If a digital platform holds personal information about an individual that was collected for a particular purpose, then the accepted norm is that the digital platform must not use or disclose the information for another purpose unless the individual has consented to the use or disclosure. Yet we harvest more personal information with every update.
Pre-emptive legislative intervention combating current and future privacy invasive AI capabilities become essential in correcting the asymmetry of power held by large scale digital platform companies for example.
The recent massacre in New Zealand that was live streamed to Facebook in it’s seventeen-minute entirety has been the direct cause of the controversial and fast tracked, Sharing of Abhorrent Violent Material Bill 2019.
This bill introduces new laws threatening platform executives with three years’ imprisonment if material deemed abhorrent is not removed “Expeditiously.”
Currently, digital platforms require the user to consent to often ambiguous privacy policies in order to use the software, such as making an account on Facebook.
Once the user accepts the aforementioned, ‘take-it-or-leave-it’ agreements, the personal information is then collected at will, such as locations and biometric data. It is then shared with multiple third parties at any time, unbeknown to the user.
An example can be seen in the Facebook auto-tag ‘scandal ‘whereby biometric AI software scanned uploaded photos and presented the name and profile of a user in the photo. This was recognised as a privacy breach and the AI function was made illegal by multiple countries world-wide.
Personal information has been a threshold legal issue for some time in the world of privacy. In combatting the methods of privacy invasion consenting to the use of this information is essential.
Protections may include that consent to sharing personal information must not be forced or as a condition of use for a digital platform, beyond what is absolutely necessary for its regular, user intended use.
This would mean if an individual wanted to join a digital platform, purely to privately message friends, they could do so, without the platform tracking their location data and sharing it with third parties. If an individual wanted to use a digital platform for shopping and wanted to receive targeted advertisements, they could opt in and do so.
As we live in a world where undersea fibre optic cables trace the legacy of imperial trading routes where people access information via search engines rather than libraries, regulations must also adapt with modern times and have the fluidity and breadth to monitor and combat negative impacts from technology such as AI.
The importance of preventative legislation is paramount as the speed of AI data collection in unparalleled. However in Islamic nations it is suggested that a global lead should be taken to ensure that the secular development of AI is calibrated to the teachings of Islam .
In a globalised economy, no matter how laudable the theological directive may be, the protection of privacy rights to the standard by which we understand Islamic thought in respect of privacy may be incompatible.
Connor William Harrison is a student of law in Queensland.