Will increasing regulation stop businesses using AI?

October 28, 2022 | 4 minute read
John Menhinick
Senior Director, Product Management
Text Size 100%:

The Context

We all know it’s coming. In fact, it’s already here and there’s much more to come. Legislation and policy focusing specifically on the use of AI is emerging globally. Examples include AI Act (EU), California Code of Regulations, CCPA, Stop Discrimination by Algorithms Act, AI Bill of Rights, Algorithmic Accountability Act (US) and many others across the world.

As individuals, we should be encouraged by this. Our privacy and our freedoms are being monitored and protected. Governments across the world will insist on high standards, particularly for high-risk AI solutions. All good. As businesses however, things are different. Brands want to provide AI solutions to their customers that are compliant of course, but are the demands of legislation becoming so onerous that they may make developing and managing AI solutions cost prohibitive? And, with penalties now being punitive for non-compliance, is it worth the risk? ‘Is the juice worth the squeeze?’ I hear them cry. Further, when it comes to using vendor-supplied AI services, can we trust that these services are themselves compliant? Should clients just focus heuristics and general process improvements internally?

An understandable viewpoint. However, at a time when businesses are struggling to grow, protect existing income streams and find efficiencies in their operating models, it would be a real loss if they could not derive the performance benefits that AI can bring to bear.

The pathway forwards

Despite the inherent difficulties in ensuring compliance in an increasingly legislated world of AI, the impact of these regulations can be viewed positively from a business perspective. Indeed, aspects of these such as explainability, may provide the opportunity for businesses to engender trust among customers and potential customers and a genuine source of differentiation.

‘Companies that can develop AI algorithms with stronger explanatory capabilities will be in a better position to win the trust of consumers and regulators.’     Source: Harvard Business Review.

Whilst there are sector, risk and geographic differences, legislation is essentially aiming to focus on the same key factors:

Factor

Description

Fairness

Appropriate, fair, unbiased

Accuracy

Accurate interpretation, insight, classification, prediction, guidance

Robustness

Tested, validated, consistent/reliable reproduction of results

Security

Compliant with highest global security standards particularly where PII is involved

Transparency

Ability to explain the process for AI recommendations from data ingestion to output

Data quality

Confirmation on the type and provenance of all data used in the AI solution

Controllability

Ability to turn the solution off/on as needed and revert to manual processes where appropriate

In view of the above, standard ‘frameworks’ are being created for use by client and vendor organisations to make achieving compliance more straight forward. For example, in Germany, the Bundesamt für Sicherheit in der Informationstechnik (The Federal Office for Information Security,) is creating a framework to assist in compliance audits. Businesses will be able to use frameworks like this to determine how compliant they are across a range of factors (including those listed above).

Technology vendors are also lobbying the organisations creating such frameworks to ensure they are as simple to understand and practical to implement as possible. Rest assured then; work is being done behind the scenes to make it easier for business customers to put the steps into place to use AI in a compliant manner.

In addition, vendors are working to ensure they continue to be compliant as new legislation emerges to give customers the reassurance they need that the tools and services they are using are based upon technology and processes are themselves compliant.

Embedded AI, that is AI incorporated into the business software you use to run your business, is a simple worry-free way to leverage AI via the software you organisation uses every day to execute its core workflows, in a compliant manner.

Conclusion

Whilst Business may fear the additional obligations and investment requirements placed upon them by new AI legislation and tighter industry regulations, help is at hand.

AI Frameworks are being created that will help organisations develop and manage AI solutions in a compliant manner. Added to this, technology vendors offer compliant solutions either as embedded AI or via AI platforms, tools and services that enable them to build solutions that are robust, transparent, and secure.

Ironically, the tightening of legislation and regulations may actually grow the AI market. If businesses know that the AI solutions they are using are compliant and protect the rights of individuals, they can be confident that these solutions will be brand-enhancing and not increase their risk profile. The future for AI in business is bright then and one where they can continue to leverage AI at a time when they need it most.

For more information on Oracle AI see Oracle Artificial Intelligence.

John Menhinick

Senior Director, Product Management


Previous Post

OCI Speech provides automatic transcription and captioning for customers of the Kaltura video platform

Jay Hardikar | 2 min read

Next Post


Custom models and text translation come to OCI Language

Pratap Kondamari | 5 min read