Appageddon: How to select the best AI applications for marketing and comms

Whenever there is technological innovation, marketing and communications are some of the first domains to pivot and adapt — and that’s the case with the willingness of this sector to adopt new artificial intelligence tools.

But right now, the AI app market looks like a maelstrom.

Obviously OpenAI’s ChatGPT app is a market leader, hitting half a million downloads in just six days, showing the same surging demand that the web application experienced when it launched late last year.

Yet even before the official OpenAI app launched, there were dozens of ChatGPT lookalikes, as well as third-party apps balancing carefully on access to OpenAI large language model, either legitimately through the API or in some complicated sticky tape arrangement predating the official release.

In the first three months alone, 160 new apps were released boasting AI capabilities, according to tracking platform Apptopia. Expect that number to double for the quarter just closed.

Downloads for AI apps were up 1500% and the money being made is also impressive.

While ChatGPT 3.5 is still free, many apps are making money leveraging AI-enthusiasm by lifting caps on the word limits that can be used in requests, curating prompts or promising greater functionality.

Genie, an app that offers each of these features, is estimated to have earned $1.5 million just in March, which would put it in the ranks of the top 1000 app earners.

For corporates looking to engage their marketing or communications teams with AI, the explosion of easy access to apps — some of them questionable — creates challenges.

Unlike traditional software, which has to go through some form of IT or at least financial gatekeeping steps before licences are rolled out, AI apps allow any dabbler to start waxing lyrical at work, with all the governance, privacy, security, and data risks that go with phone-based applications.

This explosion poses an immediate reputation risk — are your employees plugging details of a future market-sensitive release for the ASX into a phone-based tool to write the update?

Are they sharing creative openly on a tool where a client competitor can see the work?

Or is their content being dulled thanks to ChatGPT’s distinctive tone?

In an ideal world, corporate marketing and comms teams would only work within the walls of legitimate, tried and tested, and properly licensed tools.

But while we wait for OpenAI to finally release an enterprise grade ChatGPT for business, here’s a quick guide to navigating the AI app market.

1. Choose verified and supported apps

Not everything that adds the term AI has anything to do with the well-trained large language models that have seized headlines. With so many apps tagging themselves as ‘ChatGPT,’ it’s critical to ascertain the authenticity of the app and its developer. Now that OpenAI has released its official app this should be easier, but be wary of other tools that are marketed under other lookalike names, such as the AI app Ernie, created by Chinese tech powerhouse Baidu, which is now planning to sue Apple over fake versions around the app store. Dive into the app’s background, check for any affiliations with reputed organisations (such as OpenAI, Microsoft and Google), and ensure it has an active support system.

2. Prioritise privacy and security

The risk of choosing an unverified app isn’t just that you are splurging your cash on a counterfeit — scammers have seen the demand for AI apps and are taking advantage of the speed to market to sneak in apps with malware, data-theft tools or so-called dark pattern user design, which can cost unsuspecting users money. Cybersecurity experts Sophos in April reported the rise of what it calls ‘fleeceware’ — and included Genie in that list — criticising apps that charge premium prices for what ChatGPT offers for free.

Beware any app that has poorly described data handling and security settings, or which doesn’t disclose what data it collects and where it is stored or shared. Read the fine print, seek out user reviews, and be meticulous.

3. Beware app bloat

Apps often market themselves as the expert in one particular skill — translating text into multiple languages with AI, or creating social posts, being a better blog writer, or jazzing up your job ads. But once you start to add a monthly subscription here and a one-off download cost there, you can quickly erode your tech budget. Be laser-focused on your requirements and resist the temptation to hoard a multitude of apps. Sometimes, one solid foundation, like ChatGPT, can be leveraged for multiple use-cases, saving you the hassle of wrangling with an army of apps.

4. Avoid procurement challenges

Start-up app makers often don’t understand the hoops corporates have to go through to procure new software. Opt for providers who can proficiently engage with account teams, provide troubleshooting support, and collaborate with governance teams for data storage requirements.

5. Fail fast, implement thoughtfully

The pace of AI development is extraordinary, and corporates are rightly scrambling to keep up. But the approach taken for ordinary software integration might not work if you are looking to leverage the technology any time this year. Having a team or individual who is able to rapidly road test and experiment with shortlisted apps can cut down the time to implementation, but working to the ‘fail fast, learn faster, implement thoughtfully’ mantra. This nimble approach saves time, resources, and prevents the risk of going all-in on an ill-suited solution.

It is rare that we can say that the answer to a marketing and communications solution is readily available — and it is a surfeit of options that is the challenge, not a lack of opportunity.

But teams looking to leverage AI in their work still need to be discerning, to scrutinise the fine print and take an agile approach. Embrace authenticity, protect your privacy, stay lean, sail through procurement, and be ready to experiment.