AI: Can you avoid the risks it carries?

Are there risks to AI? Absolutely. There are end-of-the-world predictions about the use of IA. For a business, many of the risks are a bit less extreme, but they are also very real. For example, in the area of content creation. There are a variety of risks that you open yourself up to. One of the key ones is the trustworthiness of the content created. You rely on generative AI to create an accurate explanation or description of a topic, event, thing, or idea, However, can you, in fact, completely rely on that? The answer is probably a qualified no. The level of “qualified” depends on a variety of factors. Your AI generated content is only as good as its sources, and that can create real questions for readers. Also, an organization using AI to create any type of video, text, image, or audio content needs to be concerned that it may include proprietary information that you need permission to use. Could material created by generative AI suddenly veer off into copyright infringement?

AI is also being used in areas such as recruitment. However, there has been research suggesting that bias can sneak into AI decisions as a result of the source data the tools are using. Bias is a concern not limited to the one example of recruitment. It can have consequences in areas where AI is making marketing decisions, and can taint medical and legal recommendations AI might provide. As a result AI cannot go “unmonitored.” Review by humans and other tools is a best practice that is needed to improve accuracy and trustworthiness. This, in turn, may cut into the efficiencies that are perceived to be created by AI. Also, a lot of AI–Chat GPT to just take one example–isn’t going to necessarily incorporate consideration of regulation and compliance requirements. Many countries, individual States in the US, and US federal agencies are implementing data security regulations that are designed to protect the Personal Information of individuals. In many cases violations include civil penalties. In the case of the European Union’s General Data Protection Regulation, fines are significant.

If you are considering stepping into AI, your MSP can provide guidance. Our recent list bears repeating: Eight ways an MSP can help you approach an AI solution.

Step one: Assess potential uses of AI. Your MSP should have a solid understanding of your entire business and how AI might contribute. They can help you start with small steps and move from there.

Step two: Understand your KPIs and organizational goals, from the top down. Before jumping off and adopting AI just because it is there, evaluate your KPI’s. Where do you perceive you need a boost?

Step three: Propose a possible range of AI solutions. An MSP will be knowledgeable about the variety of applications out there and lead you to select those most appropriate for your goals. Remember, they should be directed toward assisting KPI improvement.

Step four: Estimate the solution’s ROI. Remember, measurement is important. And you can not do everything. So identify each potential AI solution’s ROI. As mentioned above, AI isn’t just a trendy tool to adopt just “because.”

Step five: Ensure compliance: For example HIPAA, PCI. HITRUST. ISO27001, SOC1, SOC2. AI is a powerful and potentially intrusive tool. Compliance is critical.

Step six: Implement the solution. An MSP can implement the solution for you. Most business owners do not have the resources available for what can be a time-intensive project.

Step seven : Manage tool-related risks. As noted, there are best practices. Monitor to ensure your outcomes with AI are accurate, trustworthy, defensible, transparent and meet regulations.

AI Is All Around Us

IT seems it is virtually impossible to avoid hearing about Artificial Intelligence (AI). Ever since ChatGPT hit the market, AI has become a never ending source of news, articles, advertisements, and lots of gloom. Artificial intelligence isn’t exactly new–the term goes back to the mid-1950s. Artificial Intelligence is a broad term and encompasses a few different subsets of processes. Generally, it refers to machines or computers doing things that we consider a skill limited to human intelligence. What has caught the public eye is what is labeled “generative AI”. Generative AI (e.g ChatGPT) refers to the AI tools that can create content, music, images, code and voice. One of the reasons generative AI is so widespread in its applications is that it doesn’t require coding skills for a layperson to use it, instead the user can instruct the tool to create content by using natural language.

Questions about generative AI – The media has certainly been filled with concern about AI and has raised many questions about areas where we potentially interact with it. How do we know the content we are reading is accurate and can be trusted to have come from reliable sources that have been vetted for accuracy? Can it be used to create misleading information that could misdirect our understanding of social, political, cultural, legal and other issues that affect the well-being of society? Others worry it could displace whole sectors of human labor. These are heavy questions best left to another venue of discussion.

Where is the average person interacting with AI?

We interact with artificial intelligence everyday in places we probably never give much thought. Those recommendations for purchases that you see on every shopping website you visit? That is done by a form of AI known as machine learning. Your thermostat that turns the heat up and down by learning when you usually leave the house? The refrigerator that makes your shopping list? Those are both examples of machine learning as well.

If you use a Managed Service Provider or a Managed Security Services Provider, AI is a new line of higher quality defense against cybercrime that they may be using to protect you. One of the greatest risks a business faces is a breach of its data by cybercriminals: malware, ransomware, et.al. and the tricks being used keep increasing in sophistication. Ransomware is particularly insidious. It can seize your data and hold it hostage for a ransom of currency, crypto or traditional. Once attacked there are very few alternatives to submitting to the ransom request. AI can help MSPs respond faster to specific threats and concerns, and assist in diagnosis and troubleshooting. Also, as every SMB knows, 24/7 monitoring and support is a critical service that an MSP can provide far more efficiently than a company can do in-house. (This has to do with the benefits of economies of scale.) AI can improve 24/7 monitoring because AI can evaluate an enormous amount of data, far faster than humans, and likely identify problems before they affect your business.

AI: Of any value to an SMB?

Artificial Intelligence (AI) has been all the media rage in the past year. Specifically, it isn’t AI in general, but a specific category of AI known as generative AI. This AI is capable of creating content, such as text, images, audio and similar data. Examples of generative AI tools can create content, music, image code, and voice. What this can include are documents that are used for marketing and other content on websites, as well as images, video, and audio. What made generative AI more widespread are the tools that use natural language to utilize them. It doesn;t necessarily require expertise in coding anymore. The generative AI tool that hit the news and has everyone curious about this development is ChatGPT. This allows any user to create conversations, answer text, and similar “written responses.” ChatGPT and similar tools are available to almost anyone.

Of course if you follow the news, there is much excitement about the potential of generative AI. It may be used to facilitate faster customer service, help attorneys evaluate large quantities of legal documents and propose new approaches to cases, medical professionals diagnose, and on and on. It also raises lots of concerns. How does one know that the content created by generative AI is accurate and can be trusted? Can it be used to create misleading information, such as deceptive statements that could alter someone’s understanding of a political, cultural or medical issue. And there are others who worry it could displace whole categories of human workers, but that question isn’t our worry here.

One place where you may benefit from the use of AI is via your Managed Service Provider. Many industries can benefit from the judicious use of AI; legal, medical, architects, etc. and the MSP world is no exception. In particular, AI may be another line of defense in cyber security. Clearly, one of the greatest risks any business faces is cybercrime. Malware, data breaches, ransomware–they all are a business-owner’s nightmare. Ransomware can hold your data hostage. Once attacked, there are very few alternatives to submitting to the ransom request unless you have solid, uncorrupted backups.

AI can help MSPs respond faster to specific threats and concerns, and assist in diagnosis and troubleshooting. Also, as every SMB knows, 24/7 monitoring and support is a critical service that an MSP can provide far more efficiently than a company can do in-house. (This has to do with the benefits of economies of scale.) AI can improve 24/7 monitoring because AI can evaluate an enormous amount of data, far faster than humans, and likely identify problems before they become business effecting.

One way a small business can utilize AI? Marketing and sales. A lot of an SMB’s digital marketing tasks can be time-consuming and easily automated. Drip email campaigns, website visitor tracking, understanding where each customer exists in the sales funnel at any given moment, and other digital tools that increase customer engagement and drive sales are an excellent introduction to AI as a marketing tool. These tools both free up sales and marketing staff for other more complex tasks and improve customer engagement. These tools that can be easily deployed by SMBs are an excellent introduction to AI. Talk to a trusted MSP for guidance. You don’t have to go it alone.