Government Agencies and Corporations Are Building Frameworks for Ethical AI Use
In April, the European Commission introduced ethical AI guidelines for the accountability of Artificial Intelligence in tech. This is just the latest in a long line of political and corporate organizations preparing to implement strategies for the future of AI use.
Citing problems in decision making and accountability, there are calls for increased accountability and an emphasis on ethical considerations as the world moves forward with new emerging AI technology.
As we saw with GDPR, tech regulations made around the world can very quickly have profound effects on businesses. US companies are projected to spend a total of nearly $42 billion on compliance with the EU-based law.
AI is becoming a staple of modern business as more and more companies recognize its value. From automation to in-depth analytics, AI will become central to the transformation roadmaps of many SMBs as comprehensive options for automation become more affordable for smaller organizations.
The implications of these guidelines on how AI is used and implemented is the question on everybody’s lips. Countries with leading tech sectors like Canada and Japan have already laid out extensive plans for AI that include principles of its ethical use.
The US government has begun exploratory looks into ethical consideration for AI, and earlier this year stated its objective for growing the AI tech industry. Corporations have also got in on the act, though Google’s attempt at forming an ethics panel on the matter last spring was a notable disaster.
AI in Tech: The Benefits of Automation
The question of ethics and accountability is particularly pertinent for AI in business. AI is most commonly applied in mundane, manual tasks. These are often low value, routine manual tasks which take employees away from more valuable tasks. In many of them, humans may oversee the process, but usually, limited supervision is required. AI is commonly leveraged for the benefits it brings in applications including:
- Cyber threats: Algorithms track behavior, detect anomalies in patterns, and operate outside of business hours. AI can handle the daily routine IT tasks, automate data backup, and secure data from hacks. Specialists can instead focus on higher-level initiatives and more complicated projects requiring a human touch.
- Process automation: Such as data entry, collation, or indexing. Employees spend less time filing and more time building relationships or developing ideas to propel the business to success.
- Ongoing analytics: Both for studying customers and analyzing aspects of business operations. Currently, its use here enables a better understanding of the customer experience and how employees work.
The rapid development of AI systems among large organizations and SMBs has led to increased conern. Aside from its hiccup earlier in the year, Google has outlined a reasonable list of goals for ethical use of AI which bears semblance to themes common in other guidelines. The list makes considerations for:
- Social benefits: Respecting cultural, social, and legal norms
- Bias: Preventing algorithms and databases from unfairly reflecting biases related to characteristics such as race, gender, and nationality
- Accountability: Provide opportunities for feedback and appeal. Make AI tech subject to human direction and control
- Privacy: Create AI tech with privacy safeguards and appropriate transparency
In their Ethics Guidelines for Trustworthy AI, the EU has laid out seven similar essential requirements which it hopes will become the foundation for an ethical guideline to will shape the industry globally.
It comes hot on the heels of last year’s GDPR, which has so far found popularity by framing data privacy as a fundamental right of citizens. These ethics guidelines have similarly sought to frame fundamental human principles at being the center of AI systems.
The key elements emphasize human oversight, data privacy, and accountability—noting that as AI technologies become more prevelant, there should be a concerted effort to take into account major public concerns.
These guidelines are important because they shed light on the race for supremacy in AI. Europe is notably lagging behind China and the US in terms of tech investment, but has found global support for the introduction of guidelines in other areas such as emissions laws and food standards.
This new push can be seen as an attempt to lead the way on a topic that is quickly gathering steam. By setting themselves apart, they can gain the upper hand in leading the discussion worldwide of ethics in the AI industry.
Such an approach is neither unwarranted nor disingenuous; the priorities of AI strategies around the world vary wildly in their objectives. The US, for example, is currently leading the opposition to a worldwide ban on killer robots. China, on the other hand, has primarily focused on its controversial social credit system which looks to take advantage of its total access to citizen data.
Both emphasize the use of unmonitored AI—that is, fully automated technology with little to no human oversight that is making decisions on how to deal with people. The core of European concern is that such technology might eventually enable businesses—and perhaps governments—to relinquish responsibility for making decisions or use AI for controversial purposes.
How Will US Businesses Be Affected?
Of course, these guidelines are not enshrined in law, as with most approaches to AI by governing bodies. However, they do serve as a foundation for future conversation on the topic and could form the structure of future regulation, with testing of the guidelines with organizations beginning in 2020.
With California’s Consumer Privacy Act—“GDPR in the US”—taking effect next year, it’s not unreasonable to assume that future tech laws will also take inspiration from across the pond. The US, along with 41 other countries, has already endorsed a recent move by the Organisation for Economic Co-operation and Development (OECD) to establish non-binding principles of AI.
After the preperations and costs that had to be made after GDPR came into effect, US businesses will be looking with curiosity to see how these guidelines will affect the organizations down the line.
Considering the bad press that some large US companies have already received, and in spite of solo efforts to make in-roads with regards to AI technology, businesses might find themselvesadapting to lawmakers regulations sooner rather than later.
This could essentially mean that, as with GDPR, organizations would be well-advised to assess how best to prepare for potential bumps in the road ahead. The use of big data, for example, utilized en-masse to better the capabilities of AI tech, is a practice which may find itself in the firing line of data transparency concerns.
The Future of AI and Automation
AI in tech delivers many advantages to businesses by streamlining processes and eliminating the chore of low-level, repetitive tasks. However, these same benefits could quickly become problematic because of the means to achieve them.
This is particularly the case with practices such as the collection of vast amounts of customer data to improve AI-human response. Will consumers still be comfortable with this in the near-future? If not, SMBs which are employing extensive AI strategies into their organizations will have to seriously consider a plan to anticipate upcoming challenges that the use of AI might pose with tighter ethical regulations.
There’s no denying how AI helps businesses, and its use will continue to grow in the near future. However, we’ve already gotten a taste of how people react to advanced technology when it concerns privacy.
Ethical AI practices are receiving similar scrutiny as data protection and privacy have, so it’s a good idea for SMB decision makers to be aware of the considerations they’re more than likely going to have to make as the technology becomes more sophisticated.
Want to Learn More?
Artificial intelligence is a powerful tool that organizations can leverage to increase productivity and efficiency. Read more about intelligent process automation and reach out to start a conversation today.