Trinity College Dublin’s ADAPT Centre recently hosted a conference to advance and develop standards in AI. More than 150 experts from 22 different countries and firms such as Microsoft, IBM, Google, Huawei, and Fujitsu attended the five-day event.
Delegates formulated policy in the areas of foundational AI standards, big data, AI trustworthiness, use cases, applications, computational approaches and ethical and societal concerns.
AI must advance in a way that builds user trust and is accountable and acceptable to society, says Prof Dave Lewis of the ADAPT Centre.
“By working to standardise AI, we are enabling more reliable, consistent and efficient assessment methods and tools for users, purchasers and future regulators of AI,” he said.
Critical factors
“Government and international bodies around the world are highlighting these as critical factors in building trustworthy AI.
"Being involved in international AI standards will strongly position Irish researchers and companies to exploit new commercial opportunities that arise from its adoption.”
And top-ten firm Mason Hayes & Curran says that Irish AI-specific legislation is likely to almost exclusively derive from future EU ethical and legal developments.
An MH&C briefing note points out that AI products and services already on the market are regulated by legislation such as the GDPR, the Data Protection Act 2018, the Copyright and Related Rights Act 2000, and the European Union (Protection of Trade Secrets) Regulations 2018.
However, there is a lot more to come from the EU, according to the firm’s AI expert Brian McElligott.
He believes that the EU wants to build an AI regulatory environment in its own image – one based on a set of fundamental values, complemented by a strong and balanced regulatory framework.
Trustworthy
Last December, the AI High Level Expect Group delivered its draft ethics guidelines for trustworthy AI.
The guidelines focus on “human-centric and trustworthy AI that produces products that operate in a traceable and accountable manner, and are based on a principle of ethics by design,” according to MH&C.
The EU’s view is that the ethical dimension of AI is not a luxury feature or an add-on, but that “It needs to be an integral part of AI development. By striving towards human-centric AI based on trust, we safeguard the respect for our core societal values and carve out a distinctive trademark for Europe and its industry as a leader in cutting-edge AI that can be trusted throughout the world.”
Canny
McElligott believes that the long game being played by the EU is a canny approach that could see it trump its US and Chinese competitors in the long run, particularly in a world of consumers who are becoming very privacy-savvy and are conscious of the downsides of owning and using products and apps that require being fed large amounts of data, including personal data.
In such a world, trust, and not speed, could be the ultimate driver of innovation, he says.
Brexit instability could cast Ireland in a very favourable light for its burgeoning AI sector.
Legal concepts
“From a technology contracting perspective, our legal concepts are recognised and understood by most foreign investors, including US multi-nationals.
“The stability offered by Ireland being intrinsically tied into the future of the EU AI legal regime is another positive.
“In addition to our 12.5% rate of corporation tax on trading profits, which applies to companies that actively exploit IP [intellectual property] through Ireland, the Irish tax regime also offers a knowledge-development box regime.
"This regime provides a highly attractive tax rate of 6.25% for income generated from commercialising patents and copyrighted software. This is particularly helpful for AI entities seeking to qualify for relief,” McElligott points out.
Fast track
Ireland’s first Masters’ degree in AI was recently launched in response to a growing demand by industry for related skills in Ireland. The University of Limerick programme includes a fast-track introductory course, developed in collaboration with the Irish Centre for High-End Computing.
Companies involved in developing the programme to date include: Accenture, Advanced Metadata, Analog, Arvato, Citibank, Ericsson, Fujitsu, GM, Google, IBM and Microsoft.
Strong signal
“The speed with which this course was designed and developed on foot of that industry demand is impressive, and sends a strong signal to international partners that Ireland is willing to take specific action to support the AI sector,” says McElligott.
A PwC report concludes that AI will boost Ireland’s GDP in 2030 by 11.6% to €48 billion.
“The Irish AI ecosystem is strong and growing from a collaboration, research, education and economic perspective.
“So far, all the signs are positive that we can live up to the projected growth rates and our growing reputation as the AI island,” McElligott concludes.