Is the push into AI robbing us of our human rights?

According to Gartner Group between 2018 and 2019, organizations that have deployed artificial intelligence (AI) grew from 4% to 14% and this trend is expected to rapidly increase over the next year.  But one may argue that perhaps we are expanding too fast into this field without looking at the downsides it may pose to our basic human rights to live in freedom and safety and measures that can be taken to secure us from the threats this technology presents in today’s world.

If you have a pulse you have heard the phrase “fake news” plastered over every media platform in the world.  But this idea of fake news or fake content takes a more sinister and potentially deadly twist when applied to AI.  And so encompassing is this notion that experts attending last week’s Wall Street Journal’s Pro Cybersecurity Executive Forum in NYC consider fake content perhaps the leading threat AI will pose in the near future. A threat that overlaps and impinges on the  Universal Declaration of Human Rights Article 3  that states “Everyone has the right to life, liberty and security of person”  and that no one should be deprived of such rights without good reason.   

Safety

One of the scenarios presented at the Forum was that of a cyberattack on AI programming in the auto industry.  For this article let’s take a look at the potential for a hacker to break into the long-awaited Tesla driverless car.

Musk has been teasing since 2017 that his vehicle with full self-driving capability will drive cross country but if and when that day comes what would happen if a hacker changed signs in the program to give the software faulty information that can cause a deadly accident.  Or extending out that scenario to other unrelated fields, what could happen if a hacker altered data monitoring for AI-assisted surgery?  What measures are in place to guard against such an attack or to warn of an impending one?

But cyberattacks are not the only concern with safety when dealing with AI.  There is the “non-human” element as well.  Currently, researchers in Australia are developing an AI-powered scribe, one that is able to chart patient-physician encounters in real-time.  Those researchers from Macquarie University are looking to obtain tens of thousands of actual physician-patient conversations in order for the AI to learn.  The problem is there are so many slangs and accents and ways in which people speak that it may create a situation whereby vital information is incorrectly entered due to a lack of understanding.  This paves the way for records to be transposed and medical procedures and emergency-related information to be incorrectly entered with potentially life-threatening results.  The Joint Commission’s mission is to continuously improve health care for the public, by evaluating health care organizations and inspiring them to excel in providing safe and effective care of the highest quality and value.  To this end, they have documented and suggested best practices in various fields such as disinfection, pharmaceuticals and smart alarms to reduce errors. I would expect to see the field of AI to be on their radar in the near future to ensure there are measures in place to keep AI as a needed resource for medical advancement without the potential negative consequences of rolling it out too early. 

Right to Privacy

But the threats with AI are not just situated in safety issues but in basic qualified human rights issues as well.  We as humans have a right to privacy and to be judged not on our past transgressions but on our current conduct.  However, it is impossible to read the news without hearing about face recognition being used in stores to identify past shoplifters or in the medical field where conversations are needed to be recorded between doctors and patients. Although AI has tremendous benefits to society at what cost does this come?

Retail Face Recognition

In the United States alone more than 800,000 inmates are released each year.  Every one of these men and women have paid their debt to society and are looking for the opportunity to become productive citizens again without the stigma of their past.  Yet, with the use of face recognition in stores, these same people are being watched and judged because of their past crimes. And who exactly are the shoplifters in retail stores? 

Statistics have shown that on any given day in America alone there are over 550,000 shoplifting incidents.  550,000 a day!  That is an astounding number.  And to add to that it is believed that 1 in 11 people shoplift at some point in their life and that most of them are normally law-abiding citizens where the temptation was too great.  Keeping face recognition on this large of a group will become unmanageable and everyone, including you and your neighbor, may eventually be in this face recognition software and be judged by those in retail every time you enter the store.  Why do I say you may be part of this record-keeping?  Simply because there are no rules governing this intrusion into privacy.  Faces being kept are not those who have been convicted of shoplifting but rather anyone who has been accused of such a crime whether or not they were later proven to be innocent.  This means every one of us is a target.  And one possible group of particular concern is the population under the age of 18.   25% of the shoplifting population are teens.  By law teens, criminal records can only be read by the teen themselves, the victim, a police officer, or a person in the legal or justice sector.  But remember with retail face recognition a person does not need to be convicted to fall into this watchful and potentially punitive state.  The person just needs to be accused or believed to have taken something.  So if a teen purchases an item and the tag is left on does this person fall into the face recognition nightmare.  This begs the question if retail outlets will be keeping a file on juveniles which is in direct opposition to the legal rights of a child.

Conclusion

AI will bring about unheard of benefits across every sector in the world including those tasks with automation, medical, natural disaster responses and more.  But as great of a benefit that will be derived from artificial intelligence, there is a greater need to ensure that the threats are fully understood and measures in place to address them before it is utilized in a widespread fashion.  Humans before machines and artificial intelligence for the benefit of humanity.