In his latest Development & Insight, Bank of Ireland’s head of Tech, Media and Telecoms Paul Swift dissects for businesses the difference between AI and GenAI.
“SMEs can benefit from engaging with AI specialists and solution providers with knowledge of their sector”
ChatGPT recently celebrated its second birthday and continues to add to the ‘noise’ associated with all things related to AI. We continue to hear from customers perturbed by the omnipresent artificial intelligence (AI) chatter, wondering how to integrate the technology in their businesses or to better understand if they even need it at all.
AI: Refers to how systems or machines are trained to mimic human intelligence to perform tasks. Machine learning is used to train the machines on data and on various training models. This can be applied in a range of different business applications such as predictive analytics to identify trends in sales, automating repetitive tasks or optimising supply chain routes/resource allocation. In short, the bigger the problem and the more data that can inform the training model, AI is more appropriate to help solve the problem.
GenAI: On the other hand, generatative AI is a subset of AI that creates content based on patterns it has learned having been trained on existing data. It generates outputs that mimics the style or characteristics of the data on which it was trained. In the time since its launch ChatGPT has helped transform how businesses have gone about generating content for marketing copy, designing packaging ideas or design prototyping and creating personalised customer experiences.
Which do you choose: AI or GenAI?
It depends on:
- What problem are you trying to solve:
- If you are trying to improve efficiency: AI.
- If it is a creative or innovative task: GenAI.
- Try, test, iterate:
- Start with small projects to understand what works best in terms of your objectives.
- Seek expert advice: SMEs can benefit from engaging with AI specialists and solution providers with knowledge of their sector.
In summary, while there is no right or wrong answer, AI models will over time become democratised and more accessible to businesses in every sector. However as with all challenges faced by businesses, they are invariably situation dependent.
Hence which model a company chooses will come down to their own preferences. Looking ahead, according to Gartner more than 80% of enterprises will have used GenAI APIs (Application programming interfaces) or deployed GenAI-enabled applications by 2026. In the meantime, there are numerous tried and tested, ‘off the shelf solutions’ that can assist businesses with all types of operational, innovative and strategic challenges.
Regulation on the horizon
The Digital Operational Resilience Act (DORA) took effect from January 17th, 2025. The act contains rules that are designed to strengthen the information and communication technology (ICT) 43 security of financial entities (banks – digital and traditional), e-private equity houses, e-money/payment institutions, insurance/reinsurance, credit institutions and asset managers operating in the European Union.
The act not only imposes strict requirements on the financial sector, but it also imposes adherence to critical ICT third-party service providers. This determination as to whether the service provider is critical is based on the entity having a systemic impact on the provision of financial services; the reliance/importance the financial entities have on the service provider and whether other service providers could offer similar services (substitutability of the entity). Importantly, if the service provider is part of a group, then the entire group will be considered critical.
Financial institutions have become increasingly dependent on the services of third-party ICT providers for SaaS products and cloud infrastructure to deliver their services and offerings, but also bring with it additional risks for financial institutions and indeed for any company tapping into third-party services.
These risks were never more evident than those witnessed during the CrowdStrike outage earlier this year, which impacted banks, airlines, transport companies and private individuals. That incident emphasises the importance for businesses to continually assess their reliance and resilience in relation to third party service providers and potential for external breaches.
DORA is wide in scope and seeks to introduce a far-reaching piece of regulation that can be applied to any financial institution, whatever their size or business model. For the average consumer, it seeks to ensure the stability and security of the financial systems we rely on every day.
Through enhancing the digital resilience of financial institutions, it seeks to protect individuals’ personal financial data from cyberattack or operational disruption. It also requires financial institutions to be able to continue to operate during a cyber incident or IT failure, reducing outages. These measures will go some way to reassure consumers that their data is secure and the systems that manage their money are robust and secure.
Voice cloning and fraud
Recently, it has been reported that the voice of David Attenborough, the globally renowned broadcaster, had been cloned by fraudsters and was being used to deliver partisan news bulletins in the United States. Earlier this year, OpenAI’s chatbot, Sky, sounded eerily like the actor, Scarlett Johannson, which caused such a furore, it resulted in the chatbot being paused.
This is nothing new of course, as we have similar cases here in Ireland where for example, a fraudulent investment advertisement on social media mimicked the voice of David McCullagh of RTE along with other instances that sounded like well-known political figures or businesspeople. The worry here is that this can now happen to anyone’s voice. Fraudsters could potentially use voice cloning software to generate synthetic speech that was found to be remarkably like the targeted human voice. Barriers to voice cloning are reducing due to the accessibility of artificial intelligence (AI) applications.
These systems can not only imitate the sound of a person’s voice; but using text synthesis technology, can be deployed to automatically generate content that mimic’s the target’s style and vocabulary. This could inevitably lead to bad actors expanding their range in how they try and defraud businesses and consumers in the time ahead by cloning voices of people they do business with, or are close to the target victim, which might seem authentic at the time, until it is not.
How can businesses and consumers take the necessary precautions:
Businesses:
- Awareness and education: need for businesses to be alert to potential scams and have processes in place, similar to the pilot checklist that must be adhered to before ever parting with sensitive information or financial accounts.
- Multi-factor authentication: worryingly many SME’s still do not deploy MFA. It is highly recommended by Ireland’s National Cyber Security Centre (NCSC) to enhance a company’s security posture. This can be strengthened by using number matching which prompts a user to enter numbers from a security identify platform to approve authentication.
- Voice biometrics: deployment of text independent voice verification can be used to verify the bona fides where a person speaks a specific passphrase, enabling authentication.
Consumers:
- Call back: if there is any doubt, call back on a known or trusted number.
- Limit personal information: always need to be cautious about sharing any personal information over the phone.
- Be sceptical of urgent requests: always be wary of requests for urgent action to transfer money or seeking personal information, particularly if the request is completely unexpected.
Scale Ireland manifesto
Scale Ireland recently launched a manifesto in advance of the recent general election. The document highlights ten priorities for the next government and comes on the back of meetings held with the various political parties on the key issues facing Ireland’s scaling businesses.
Some of the more noteworthy priorities included are Scale Ireland’s ambition is to see a doubling of large Irish exporting companies while significantly increasing R+D activity across Ireland’s SMEs by 2030.
With Enterprise Ireland due to launch a new five-year strategy in early 2025, it will be interesting to see how or if the objectives of both entities dovetail. Interestingly, Scale Ireland are also calling for a new Scaling Division in Revenue, which will be tasked with supporting start-ups and scale-ups, while ensuring compliance.
This is an interesting approach to take and would go towards providing a clearer understanding of the types of challenges faced by this important sector. It would also bolster dialogue and ongoing learning and shared experience between Revenue and the sector, which can only be of benefit to all over time and help in boosting the take-up of various enterprise schemes by suitable companies.
Support for an independent and viable tech start-up eco-system that would play a role working in tandem with government and state agencies to help scaling companies succeed globally is also on the wish list. Having an independent ecosystem is seen as vital to provide support and provide a voice for scaling companies all around the country.
-
Bank of Ireland is welcoming new customers every day – funding investments, working capital and expansions across multiple sectors. To learn more, click here
-
Listen to the ThinkBusiness Podcast for business insights and inspiration. All episodes are here. You can also listen to the Podcast on:
-
Spotify
-
SoundCloud
-
Apple