The Age of AI: Practical uses and regulation

The promise of artificial intelligence (AI) has been around since World War II but it is only in the last decade or so that technology and engineering have gradually caught up with expectations. In fact, AI now seems to be on a winning streak. Having presented a brief history of AI, recent developments and potential dangers, the final chapter in this article focuses on current practical uses and the regulatory perspective.

Having mulled over a brief history of AI, and checked out the various modern approaches and recent developments, and addressed the potential dangers of AI,  the final piece in this article series focuses on current practical uses and the regulatory perspective.

Privacy concerns are being addressed with the likes of GDPR in the EU, PIPEDA* in Canada, the Australian Privacy Act and many similar pieces of legislation in various countries. The US has a mishmash of federal, state, and industry regulation. Such laws however sometimes lack to account for evolving modern technology. GDPR “has resulted in little more than millions of website pop-up privacy notices, a few fines, a giant bureaucracy, and apparently still allows ‘law enforcement’ to record all people in public and punish those who choose not to comply” (Sammy Migues, principal scientist, Synopsys). In reality, we buy and use tech from dubious provenance. Our cameras store images without our knowledge, Smart TVs and home assistants record our most intimate conversations because many lack the knowledge to understand or read the fine print. Visiting other countries requires you to bare all in the interest of security. Everything you say or write on social media or email may and will be used against you. The same GDPR touting EU has a very comprehensive “Legal Intercept” legislation (Council Resolution 96/C 329/01) that forces telecommunications providers and manufacturers to provide a permanent back door to all communications, effectively allowing Big Brother to snoop on you whenever and wherever. Throw in an AI agent in the mix and then what?

Regulation and legislation needs to be sane, realistic and balanced. In a globalised world, regulatory competition and market forces together should arrive at a happy medium half way between the free for all Orwellian implementations in China and some of the overbearing proposals in EU jurisdictions. Regulation is inevitable but posing unrealistic expectations such as explainability and expectations of perfection on a yet infant tech will move innovation elsewhere. Large corporations are weighing in on the debate and are naturally pushing for self-regulation. During the recent Davos summit, IBM’s Christopher Padilla offered suggestions on ‘precision’ regulation of AI; a so-called “sensible” hybrid approach where both government and industry come up with a set of rules that have some teeth to them. Just recently,20 IBM and Microsoft have joined up with the Vatican, the EU and United Nations to push for ethical development in the field.

The pressure for transparency in the use of AI will not go away. The Algorithmic Accountability Act21, is a bill recently introduced in the United States Senate to empower the FTC to scrutinise consumer-facing automated decision systems for bias, privacy and security risks. It will certainly not be the last.

Current uses and opportunities…

We have already touched upon several high profile uses. Real time speech to text and Real time human quality translation is a practical reality. More mundane every day presence of AI is found in predictive search. It can be very uncanny to have your mobile come up with a suggested search term after typing two or three characters. In the field of medicine, predictive capabilities of tumour detecting AI agents can match the accuracy of the best experts in the field and most of this power is available through your humble phone and augmented by the cloud.

In science, AI agents are making headway in the prediction of complex weather patterns. In astronomy, machines are taking the brunt of sifting through petabytes of data from various ground and space observatories in the search for other Earths whilst CERN22 stays ahead of hackers trying to break into the Large Hadron Collider’s (LHC) massive worldwide computing grid by using AI.

Nearer to home, toys with an uncanny ability to learn and adapt to children are being used for therapy and companionship. Cheap Drones can navigate complex and changing environments with no human input to the point of seemingly possessing a modicum of biological intelligence.

Hospital chains like Cleveland Clinic and Geisegner are making use of machine learning algorithms23 (trained on depersonalised patient records) to predict the likelihood of Sepsis on patients up to 12 hours before clinical determination. At the same time, Bayer® is planning to use third party ‘federated’ data to identify spot gene mutations24. This approach is novel both from a technical and legal perspective as it challenges the GDPR legal envelope by training its own model on data where the data resides. Such an approach is imperative to achieve any results since with rare mutations, the base dataset on which an AI can be trained is very small. Single organisations only have images from 20 or 25 patients.

In the automobile world, AI is making inroads beyond the area of autonomous vision. At a recent CES, Bosch demonstrated a virtual visor with an integrated camera based on a set of AI algorithms that can track a driver position and facial features to shade only the portion of the visor to block the sun from the driver’s eyes.

In agriculture, Deere & Co. are marketing a farm robotics application that uses computer vision to identify between healthy plants and weeds claiming savings of $30 per acre and a reduction of 90% of agrochemicals.

Insurer Travelers Cos. in California is using AI to assess property damage caused by wildfires. An image-recognition system analyses thousands of images and identifies with more than 98% accuracy which properties can be considered a total loss. According to CIO Mojgan Lefebvre, advance claim payments can be made on 90% of all wildfire total loss claims, prior to an in-person inspection.

In business, Robotic Process Automation (RPA) uses AI to automate high-volume repetitive tasks that require an element of cognitive ability. Repetitive processes such as filing submissions, populating paper works and data extraction are typical uses of such software. In side-by-side comparisons, they have been shown to exhibit greater accuracy.

The move of business software to the cloud by most major vendors from Salesforce to Oracle to SAP is accelerating the integration of AI in the product offering. From marketing to customer engagement, cloud software is creating an illusion of a cognitive interaction. Responses appear less mechanical and random with each version of software and the level of personalisation offered to today’s customers makes it very difficult to understand whether one is dealing with a human or an AI bot.

With the new AI revolution, companies must rethink how they operate and conduct business, from the boardroom to the field. As organisations continue to adopt AI, it is important for the Board, C-Suite, and business leaders to incorporate a risk and ethics framework as a key governance component of their programs.

The Age of WithTM

A world where humans work side-by-side with machines and data coincides with actions is already upon us. The “Age of WithTM” is here. And the potential is everywhere.

Deloitte is helping companies harness that power to identify unique advantages through AI and analytics to move faster with greater precision, to pinpoint truths that improve decision-making, and to create beneficial connections with customers.

Deloitte was recently named a leader by Gartner in 2020 Magic Quadrant for Data and Analytics Service Providers26, recognised for its ability to execute and completeness of vision.

At Deloitte, an AI-led approach and cross functional capabilities across business lines – including tax, risk, audit and consulting – has the potential to re-shape companies, solve their most complex challenges and change how they do business. Through its global network, Deloitte will continue to expand the reach, depth and breadth of these capabilities to provide differentiated value to clients across the globe.

Whether its data insights strategy, data governance, AI adoption journey or solutions such as Deloitte robotic process automation (RPA)25 service offerings that simplify the drudgery of repetitive office work, service providers like Deloitte enable the delivery of selective initiatives focused on trusted data-driven insights and decision making for business leaders.

If AI is not already on the agenda, now is the time for business to leverage humans with machines to drive competitive advantage, improve business operations and achieve better outcomes.

*Personal Information Protection and Electronic Documents Act


[1] A. Ng, “Andrew Ng: Why AI Is the New Electricity | Stanford Graduate School of Business,” 2017.
[2] J. McCarthy, “Arthur Samuel : Pioneer in Machine Learning.”
[3] Sebastian Schuchmann, “History of the first AI Winter - Towards Data Science,” 2019.
[4] D. E. Rumelhart, G. E. Hinton, and R. J. Williams, “Learning representations by back-propagating errors,” Nature, vol. 323, no. 6088, pp. 533–536, 1986, doi: 10.1038/323533a0.
[5] “Alvey: the betrayal of a research programme | New Scientist.”
[6] “House of Lords - AI in the UK: ready, willing and able? - Artificial Intelligence Committee.”
[7] B. Butler, “In Cloud Computing Price War Amazon and Google Drop Rates Again | CIO,” 2013.
[8] Jackson Joab, “Price war! Amazon cuts cloud costs to counter Google | Computerworld,” 2014.
[9] S. Ranger, “Microsoft, AWS and Google may have just started the next cloud computing price war | ZDNet,” 2017.
[10] A. N. Bloomberg, “Artificial Intelligence: Reality vs Hype - YouTube,” 2019. [Online]. Available: [Accessed: 23-Jan-2020].
[11] W. Knight, “Reinforcement learning,” Technology Review, vol. 120, no. 2. Massachusetts Institute of Technology, pp. 32–35, 01-Mar-2017, doi: 10.4249/scholarpedia.1448.
[12] F. Woergoetter and B. Porr, “Reinforcement learning,” Scholarpedia, vol. 3, no. 3, p. 1448, 2008, doi: 10.4249/scholarpedia.1448.
[13] Jia Deng, Wei Dong, R. Socher, Li-Jia Li, Kai Li, and Li Fei-Fei, “ImageNet: A large-scale hierarchical image database,” 2009, doi: 10.1109/cvprw.2009.5206848.
[14] “DARPA Grand Challenge - Wikipedia.”
[15] “Tesla Model 3 Driver Claims Autopilot Saved Lives In This Accident.”
[16] “Watch Tesla Model 3 On Autopilot Avoid Crash With Distracted Driver.”
[17] “Watch Tesla Model 3 Autopilot Swerve To Avoid Car Passing On Shoulder.”
[18] S. O’Kane, “Tesla hit with another lawsuit over a fatal Autopilot crash - The Verge,” 2019.
[19] J. Barnett, “The Army working on a battlefield AI ‘teammate’ for soldiers,” 2020.
[20] “Catholic Church joins IBM, Microsoft in ethical A.I. push | Fortune.”
[21] “Algorithmic Accountability Act of 2019 Bill Text.”
[22] J. Emspak, “World’s Most Powerful Particle Collider Taps AI to Expose Hack Attacks - Scientific American,” 2017.
[23] J. McCormick, “Health Systems Look to AI to Prevent Sepsis Deaths - WSJ,” 2020.
[24] Jared Council, “Bayer Looks to Emerging Technique to Overcome AI Data Challenges - WSJ,” 2020.
[25] W. Hählen and S. Kapreillian, “RPA for Tax | Deloitte Global,” Deloitte Tax Solutions.

Did you find this useful?