Artificial Intelligence And Software Testing

Gartner 2021 AI Report

I watched a Podcast, and came to know about the Gartner Report Predict 2021, which I went through to assess what it means to Software Testing. This write-up on the report claims that there are five highlight points in the report:

1. By 2025, pretrained AI models will be largely concentrated among 1% of vendors, making responsible use of AI a societal concern

What this means is that every model that’s generated by 99% of the rest of the vendors would not pretrained (or) had not been thoroughly tested (?)

2. In 2023, 20% of successful account takeover attacks will use deepfakes as part of social engineering attacks

We need to create test strategies for accounts to prevent such takeover attacks. An intersection between Security and Testing.

3. By 2024, 60% of AI providers will include harm/misuse mitigation as a part of their software

Test strategies got to figure out what is harmful use or misuse, and then figure out whether the mitigation provided is the right one. The challenge would be how to figure that out, and hopefully the developers of the AI models will give that information to the testers.

4. By 2025, 10% of governments will avoid privacy and security concerns by using synthetic populations to train AI 

Testers/Test organisations involved in validating governmental use of AI would need to watch out, and raise awareness about the ‘synthetic populations’. Though difficult, it may not be impossible.

5. By 2025, 75% of workplace conversations will be recorded and analyzed for use in adding organizational value and assessing risk

Testers who test these recording applications need to make sure that the usage is ethical. Watchdogs and panels within organisations need to be setup to make sure that the usage is ethical. This is going to be tough as employees could get fired for being the whistle-blowers, and so will be testers involved.

As the write-up points out, the outlook seems grim. As rightly mentioned in the write-up, we cannot believe what we see and hear moving forward, and have to be more diligent, especially the testers!

To discuss Software Testing viz-a-viz Artificial Intelligence, feel free to contact me.

Leave a Comment

Your email address will not be published. Required fields are marked *