Doing the proper factor with AI

Video surveillance know-how should be used responsibly and regulating know-how resembling reside facial recognition is necessary. However in appraising these applied sciences, we should be cautious to not throw away AI and its advantages with the regulatory bathwater.

All through historical past, know-how has been put to each good and dangerous use. There isn’t any higher instance of this than the private laptop. Linked to the web, the common-or-garden PC opened up the world to its customers, making communications with family and friends on the opposite facet of the world low-cost and simple.

Within the unsuitable fingers, nonetheless, the identical private laptop allowed hackers to entry inadequately secured company and authorities networks, steal delicate data and promote it to the very best bidder.

It’s proper that we frequently tackle the know-how accessible to us and intently take into account each the worth it brings and its potential for misuse. At current, there’s a particular concern throughout society about know-how that, if used irresponsibly, will infringe on our privateness and undermine civil liberties. This contains a variety of applied sciences that depend on synthetic intelligence (AI), most notably reside facial recognition (LFR). Numerous police forces throughout Europe have been utilizing LFR, however a sequence of authorized challenges have emerged round its use.

Putting a stability

LFR is one in every of many examples of AI at work in video know-how. Employed in video programs, AI can enhance picture high quality, take away noise and get rid of false alarms, guaranteeing that operators solely deal with the incidents and occasions that matter. Used this manner, AI saves time, reduces price, boosts effectivity and enhances safety. 

On the similar time, the European Fee is proposing the first-ever legal framework on AI to offer AI builders, deployers and customers with clear necessities and obligations relating to particular makes use of of AI. Central to this would be the identification of various ranges of danger related to a specific utility of AI.

Importantly, the fee additionally hopes that regulation will “strengthen uptake, funding and innovation in AI throughout the EU.” A stability is required between regulating the usage of know-how to make sure it doesn’t trample over particular person rights, whereas encouraging the innovation that provides rise to increasingly more purposes the place AI can ship actual advantages.

Accountable producers like Hanwha Techwin are used to putting these balances. To be clear, Hanwha doesn’t provide a reside facial recognition resolution in Europe. And, as a Korean NDAA-compliant producer with full management over its provide chain and a robust background in cyber safety, we’re sturdy advocates for the accountable use of video surveillance know-how.  

A proper and correct place for AI

We help correct regulation and a public debate across the accountable use of know-how resembling AI. On the similar time, nonetheless, it will be significant that the precious function that AI performs in video know-how to assist maintain our public areas operational, secure and environment friendly is recognised and inspired.

For instance, Hanwha is pioneering the usage of AI ‘at the edge’ in video cameras to resolve typical challenges that police forces face when investigating crime. Crucially, that is achieved with out impacting on civil liberties.

AI might help law enforcement officials rapidly find a lacking individual by analyzing lots of video footage and rapidly classifying photos to extract solely people who embrace attributes resembling the color of clothes worn by the person, whether or not they have been sporting a hat or carrying a bag, and even shifting in a specific path of journey. What may beforehand have taken law enforcement officials days to attain can now be performed in hours, liberating up officers to undertake extra detailed investigative work as a substitute of manually scanning video footage. AI makes this potential.

Doing the proper factor

Crucially, this use of AI doesn’t carry implications for privateness as these programs are educated to deal with objects and attributes quite than figuring out people by way of facial options.

What’s extra, AI is even known as on to guard the id of members of the general public captured on this video footage by redacting the faces of anybody besides the lacking individual and rendering them unrecognisable. Solely police and courtroom personnel and people permitted to entry the unredacted video footage could achieve this. Within the gentle of the wonderful issues that AI now makes potential, it might be a disgrace if, within the fairly correct evaluation of types of know-how that may current a risk to privateness, AI and its advantages are thrown away with the regulatory bathwater.

By Uri Guterman, Head of Product & Advertising and marketing for Hanwha Techwin Europe

For extra data, go to: www.hanwhasecurity.com