3 Questions to Ask Your AI Startup

Separating AI hype from reality is not always easy. Yes, AI capabilities are evolving quickly, but translating that into real-world business benefits can be hard. Now Deloitte is stepping up with a guide to help buyers navigate the sometimes-tall claims of AI startups.

You’ve likely been subjected to a pitch from a vendor selling an AI product or services that sounds too good to be true. It leaves you wondering: Can the AI really do that? How would that work at my business? And should I invest my company’s money to implement?

Finding answers to these questions can be difficult. But help is on the way in the form of Beena Ammanath, executive director of the Deloitte AI Institute, who has put together a handy guide to walk would-be AI purchasers through the vetting process for AI startups.

There are three important questions that customers should ask AI startups before they sign on the bottom line:

1. “Can the solution adapt?”

It’s often the case that an AI product or service that works well for one company won’t necessarily work for another, Ammanath writes. “If the product was built specifically for another client, it may need additional changes to work for you,” she writes.

At a bear minimum, the product or service should do what the company claims it can do. That is, there needs to be some actual “AI” in that AI offering, “and not just data analytics to orchestrate applications and workflows,” writes Ammanath, who is a 2020 Datanami Person to Watch.

Sometimes, an AI solution will only work for a particular industry, and not another, according to Ammanath. In addition to the ability to customize, the critical customer should also look at quality control at the vendor, how quickly the vendor brings it to market, and its ability to stay up on the latest regulations, especially around fairness, bias, discrimination, diversity, and privacy.

Mots-clés : cybersécurité, sécurité informatique, protection des données, menaces cybernétiques, veille cyber, analyse de vulnérabilités, sécurité des réseaux, cyberattaques, conformité RGPD, NIS2, DORA, PCIDSS, DEVSECOPS, eSANTE, intelligence artificielle, IA en cybersécurité, apprentissage automatique, deep learning, algorithmes de sécurité, détection des anomalies, systèmes intelligents, automatisation de la sécurité, IA pour la prévention des cyberattaques.