Partner for Personalized Outreach
Erica Salm Rench, chief operating officer at rasa.io, agrees that generative AI is “all the rage” for ease in developing outlines, ideating, and generating content, but notes that AI can also be “extraordinarily powerful to personalize content served up to each and every reader.” Her company partners with hundreds of associations to generate AI-powered newsletters via a smart newsletter solution, which allows for member personalization of newsletters and other communications.
“We also have generative AI built right into our platform,” says Salm Rench. Creating customized newsletters leads to a more personalized experience for members as well as time and resource efficiencies for association staff—plus, advertisers and association partners appreciate their promotional content being consumed by more people, according to Salm Rench.
This approach provides a deeper understanding of members’ personal preferences. “As members engage with the content within their personalized newsletters, we can take back in their topical interests, which the associations can then use for their overarching content strategies,” explains Salm Rench. “AI can aggregate data to identify trends, measure engagement, and determine the effectiveness of various communications, allowing for continuous optimization.”
Adopt AI … the Right Way
Associations that are already leveraging AI in their work have found that it must be done in a strategic manner, in alignment with the organization’s mission rather than haphazardly throwing questions or prods into an AI engine that has not been approved by association executives.
To leverage AI appropriately, “associations can start by adopting basic AI tools for content automation and analytics, or partner with AI firms to gradually scale their AI usage while learning best practices,” suggests Salm Rench. “Many of the LLM tools out there today allow folks to begin experimenting for free.”
Of course, not every communications project is suited to AI assistance. “AI should not be used for content that requires a high degree of human touch, creativity, or ethical considerations, such as sensitive communications or nuanced storytelling,” explains Salm Rench. “Cornerstone content that is meant to represent your brand in a meaningful way should be human-driven.”
Staff members should check that any AI-generated content does not contain incorrect information. AI models use algorithms to try to predict the output expected by the user, and the technology behind generative AI isn’t designed to differentiate between what’s true and what’s not true. AI also may produce “hallucinations”—new, potentially inaccurate content—by combining patterns in unexpected ways. To prevent pushing out false content, fact-checking is critical, says Cunningham.
Cunningham also cautions that generative AI can have built-in biases because the information it generates is pulled from data that it has been previously fed. To prevent biases, the data from datasets used to train AI models should be obtained from a wide range of sources so outputs are more accurate.
Cunningham encourages AI adopters to be aware of copyright laws and avoid publishing plagiarized copy. She recommends using Quetext, or a similar deep-search technology designed to analyze text, to check for plagiarism and AI-generated content.
Issues of privacy and security should also remain a priority. This “requires stringent data validation processes, robust security measures, and clear ethical guidelines,” says Salm Rench. “Create an AI usage policy for your staff, and make sure to update it regularly, as this technology is advancing rapidly.”
At ICF, Cunningham’s team developed an employee agreement for reference when using AI, with a separate agreement for freelancers. Both documents have been reviewed by the organization’s legal team. “We’re an international organization, so we have to be mindful of privacy laws and protecting people’s data,” Cunningham adds. This means that when her team feeds information into an AI model, “we keep out any identifying factors” and remove names from testimonials and other areas.
ICF staff also select the “Temporary Chat” option when using ChatGPT (and similar features in other generative AI), so that those “conversations” will not appear in the history, will not be used to generate content for others, and will not be used to improve the AI models.
Value the Human Touch
No matter how extensively an association chooses to leverage AI, the human component remains critical. “AI is a tool to enhance, not replace, human creativity and judgment,” says Salm Rench. “Prioritize human oversight in AI-driven processes and ensure data quality. Don’t rely solely on AI for all content creation and avoid neglecting ethical considerations.”
Cunningham says she is excited to watch how AI evolves in professional applications. “Don’t be afraid of what’s out there,” she says. “AI will become a part of everyday work. Get out there and experiment to see how it can help your association.”