![Story image](https://securitybrief.in/uploads/story/2024/06/26/techday_44cd5278e95b6db43f58.webp)
How the midmarket can maximise the success of their AI efforts
Half a decade ago, digital transformation was the big-ticket IT item. CIOs of enterprises of all sizes were throwing money into projects, and yet the failure rate for these projects was astronomical.
Then, as businesses and IT leaders better understood how to be strategic about transformation, the failure rate started to ease back. They are still complex projects, but there is a better understanding on how to properly map and act on them now.
Fast forward to today, and AI is on a similar trajectory. 70% of companies are reporting that AI is failing to deliver what was promised.
The midmarket is particularly vulnerable in this regard. They face the same pressures to “embrace AI” as large enterprises, particularly given that the midmarket is expected to be innovative and disruptive to the big end of town.
On the other hand, they don’t typically have the same resources to commit to building the AI strategy, and midmarket organisations that have a lot of legacy IT also face the need to undertake a transformation exercise before they can embrace AI.
As we note in our recent whitepaper: The Mid-Market and AI: How To Overcome Legacy IT And Resourcing Challenges, there are four areas in particular where the midmarket faces risk in AI:
- Operational Risks: Mid-market companies often operate with leaner resources compared to larger enterprises, which can lead to challenges when implementing AI. For example, a mid-market logistics firm might invest in an AI system to optimise delivery routes. However, due to limited data and testing, the system could fail to account for real-world variables like traffic patterns, leading to inefficiencies and potential loss of business.
- Reputational Risks: Reputational damage is a significant concern for mid-market companies, especially if AI applications result in publicised failures. Take, for instance, a mid-market financial services company that uses AI for credit scoring. If the algorithm is found to be biased against certain demographics, it could lead to public outcry and loss of customer trust.
- Regulatory Risks: Navigating the complex landscape of AI regulations can be particularly challenging for mid-market companies. A breach of data privacy laws due to an AI system’s failure to comply with GDPR, for example, could result in a company being unable to expand into Europe or even hefty fines and sanctions there.
- Cyber Security Risks: The integration of AI into business processes can introduce new cybersecurity risks. Mid-market companies may be particularly vulnerable to cyberattacks that target AI systems, such as data poisoning, where attackers manipulate the data used to train AI models, leading to incorrect or harmful outputs. Additionally, AI systems can be exploited to bypass traditional security measures, making it crucial for companies to invest in robust cybersecurity frameworks that evolve alongside their AI implementations.
Overcoming The Risks
Overcoming the risks associated with AI in the midmarket is more a matter of strategy than technology. The technology is there. Success depends on the organisation mitigating risks in three ways:
- Find an expert partner to assist: As noted on CRN, seven in ten businesses plan on seeking partner help with AI. This is for numerous reasons but ultimately comes down to having assistance with setting clear objectives, understanding the technology’s limitations, and accessing best practice guidance in the journey to adoption.
- Approaching AI as Incremental Implementation: Rather than trying to overhaul entire systems at once, midmarket companies can benefit from an incremental approach to AI adoption. Starting with smaller, manageable projects allows for learning and adjustments along the way. This method also helps in mitigating operational risks by not overwhelming limited resources.
- Cybersecurity Measures: Finally, investing in advanced cybersecurity solutions that are specifically designed to protect AI systems is key to addressing data security risks. Regularly updating these systems and training staff on cybersecurity best practices will further strengthen defences against cyber threats.
For the midmarket AI can boost productivity, underpin growth and scale. It also doesn’t need to be a complex or challenging journey. All a midmarket organisation needs to do is make sure that it is approaching AI with the right strategy and setting itself up for success first, before it starts to make those investments.