You Know More About AI in Comms Than You May Think

Artificial Intelligence (AI) might seem like an exciting new technology in 2023 thanks to the emergence of generative tools like ChatGPT, but in truth we’ve been living with it for some time. Once a product of science fiction writers and Hollywood epics – think the near-future Los Angeles in Her or Tony Stark's computer J.A.R.V.I.S. from the Marvel films – AI is firmly and productively a part of our everyday lives.

In fact, it’s probably in many more places than you realize. Machine learning algorithms, the method by which AI systems respond to user input and perform tasks, are the foundation of smart home tech like Amazon’s Alexa and Apple’s Siri voice assistants. The FaceID function you use to unlock your smartphone is AI. The Netflix recommendation engine that suggests your next series binge is also AI. All of this technology around us is powered by artificial intelligence which quickly produces seemingly intelligent outputs because it’s been trained on large amounts of data.

How the Comms World Uses AI Today

What about the world of PR and communications? It’s probably no surprise to learn that the industry has been using AI (long before the arrival of ChatGPT) in some form for a while, too.

Sentiment analysis, for example, which has been present in the comms space for over a decade, is underpinned by natural language processing (NLP) and machine learning. These two subsets of AI are used to understand the emotions expressed in text and classify them as positive, negative or neutral.

It should be noted, however, that sentiment models analyze the words people use to share their opinions, but can't account for the context in which they are used. New AI-driven methodologies have the potential to evolve and help introduce that all-important context element. At Cision we call this “Stance,” and you can learn more about it by watching our webinar, AI and the Future of Comms Teams, or reading the recap of the webinar.

So, while there has been a leap in the quality of this branch of AI, the capability of a machine to understand human language is not new. What makes it feel new, however, is that the PR and communications industry is one that remains largely text-based – so the rise of ChatGPT, built on a large language model (LLM) to interact with humans in a conversational way, has opened up possibilities for communicators to shape content with AI as a credible creative partner. As Cision Executive Director of AI Strategy Antony Cousins noted recently, the advancement and adoption of LLMs in 2023 represents “a technological leap comparable to the advent of social media or the internet itself.”

AI has long had the potential to unlock clarity and speed across all the core workflows of PR and communications professionals – and now is an exciting time in the technology’s development for comms, but even with the latest developments, AI isn’t a catch-all solution, or a replacement for people in the workforce. It lacks genuine creativity and the ability to understand nuances or “read between the lines” of a problem the way a human would.

An AI-based tool can quickly and precisely answer the specific question it is asked, but it’s also limited by the question or prompt. From that perspective, think of it as a work co-pilot that can help take care of time-consuming grunt work and enable you to focus on those higher value opportunities that matter most to your brand.

Building on AI’s Potential

At its core, AI is helping to do a lot of the work that humans would do, but at speed and scale (and in some cases, it's happening in the tools you use right now). With the help of AI and automation, many of the comms professional's routine tasks now take far less time than before.

Here are four areas you'll be familiar with and how AI raises the bar:

• Analyzing and understanding what people are saying about your brand, getting to grips with the emotional tone in a piece of text, and tracking it across multiple channels like social media, news articles or press releases, are vital for comms teams. As we discussed earlier, that's sentiment analysis, which uses Natural Language Processing (NLP) to process this information at speed and deliver a verdict.
• Reading articles to understand their relevance, then filtering and sorting them before supplying them to clients is a process AI can complete rapidly. Think of this as the AI mimicking human behavior, helping to accomplish a task that would normally be viewed as mundane and time-consuming.
• Sifting through news sources to create condensed versions of multiple articles and generating summaries of social media conversations and emerging trends can be aided with summarization AI. This is particularly useful for comms teams in the creation of media coverage briefings.
• Creating content, whether that's an article, an image for a social media post, or a press release, can now be streamlined and sped up thanks to generative AI. By leveraging technology like ChatGPT and DALLE-2, comms and PR teams can get past the blank page quickly and find new ways to engage with their target audiences.
• One other thing that hasn’t changed since we started using AI is that we still need a large amount of data to train these technologies and evaluate their output. For example, to train a model to generate press releases, you need a lot of them. “I think our earliest press release on record was delivered on horseback,” said Cousins in the recent AI webinar. “So in terms of the amount of data that we’ve got, which is where a lot of the competition in this space happens, we obviously have a lot of potential.”

"AI isn't a catch-all solution or a replacement for people in the workplace. It lacks genuine creativity and the ability to understand nuances or "read between the lines" of a problem the way a human would."

How Cision Uses AI

With our expert view to improving how comms professionals work, all of the above AI use cases are already integrated within Cision’s platforms and services. Moving forward, development and delivery of our new capabilities will be carried out responsibly, with data privacy and ethical considerations a fundamental part of how models are trained and applied.

This goes beyond being responsible just for ourselves – we understand the risks that brands (and our customers) face to equity and valuation if they're not implementing technology safely and in the right way.

“Our intention has never been to be the first getting [new AI products] out. We don’t want to take ChatGPT, [put] a user interface on it and say we’ve got a press release generator. We want to make sure that what we’re doing is right and it works for the breadth of our customers and their audiences,” said Cousins.

“We have to make sure this works for everybody and that’s a core tenet of our approach to developing responsible AI.”

Starting this year and into 2024, Cision will bring new propriety AI-powered capabilities to market:

• Risk Score: Powered by 17 NLP models, this real-time insight feature will provide clarity and context to earned media mentions, while uncovering potential risks they pose to your organisation. These signals could be indicators of racism, sexism, hate speech, fake news, humor and multiple dimensions of emotion.
• Stance: Our next generation of sentiment models, enabling automated analysis of consumer sentiment and intent in context and at scale.
• Narrative monitoring: A combination of AI technologies that can identify similar opinions in written text across social and traditional media and group them together with a single statement to explain their shared meaning. Combined with content scoring, this technology can give early warning of harmful narratives emerging about brands.
• Generative press release tool: A press release generator to create optimized press releases. We will combine this with our years of data collection on what makes an effective press release to bring industry-leading generative text capabilities.

The AI that Cision develops is vitally important to the clients and brands we serve. Many make decisions based on the guidance we provide, and as AI begins to automate these insights, they will become increasingly important for all of us.

With that in mind, we are building AI to empower and elevate our users to do more, faster and do it better while ensuring that risks are understood and transparently mitigated. Remember, this isn’t new, you’ve been using it for years, but it’s now more capable and relevant than it’s ever been.