Last year, Apple, Amazon, and Samsung had said that they are not allowing employees to use ChatGPT like tools for internal work citing privacy issues. At the same time, these companies were also developing their own tools that the employees were using, which seems like a fair way forward for a company to adopt AI.
But apart from this, companies have been extremely wary when using AI coding tools as they fear the privacy and security of their proprietary code. To tackle this, many enterprises have started adopting open source models, and fine-tuning them for their internal use, building a layer of trust for their employees.
A study by Checkmarx found that 99% of development teams use AI coding tools, even though 15% of the surveyed companies had explicitly prohibited their use. There are still hundreds of companies that are reluctant to allow developers to use AI models.
The problem is, as Santiago Valdarrama put it, “companies that don’t let their developers use AI are not gonna make it.”
Bring Your Own AI – BYOAI
Sergey Brin, co-founder and former president at Google’s parent company Alphabet said that writing code from scratch now feels really hard compared to just asking AI to do it. “I have written a little bit of code myself just for kicks, just for fun, and then sometimes I have had the AI write the code for me, which was fun,” said Brin.
He went on to add that developers are not using AI tools for coding as much as they should. This is definitely true as it was said earlier that a single developer who knows AI is better than five developers who don’t.
Companies have been experimenting with using AI tools and they have definitely witnessed a productivity boost. With tools like Cursor, Zed, and now with the latest updates to GitHub Copilot, it seems like a wasted opportunity for companies to not allow developers to use them.
Simultaneously, developers shouldn’t wait for companies to buy AI tools as well. As the conversation is already brewing around “Cursor + Claude = Junior Engineer”, it is essential for developers to upskill themselves. “I’m not sure AI is going to replace developers, but I’m pretty sure AI is going to compete directly with developers who only write code,” said Kelsey Hightower.
As for the companies, which are made up of these developers, it is essential for them to allow developers to use AI coding tools. For example, according to estimates, around 50% of the code at Microsoft is already being written by AI, which is evident in the case of Google as well according to Brin’s comment.
Which simply explains that large corporations that are thriving are using AI in their workings, which should be a wakeup call for others. When it comes to India, the adoption of AI tools within companies is still lesser compared to the US as there is very little conversation around what tools that developers are using.
What is the Issue?
As Mrinal Rai told AIM earlier that Indian enterprise companies are not particularly impressed with the new generative AI tools like Cursor or Claude, but are happy with the solutions that Indian IT is providing.
“These providers are upskilling their workforce particularly in code development leveraging many of these AI tools by taking care of the required security and compliance guardrails. It is not about catching up but leveraging these solutions to make their own workforce more efficient and productive to deliver better solutions to their clients by taking care of their security, compliance and privacy concerns,” explained Rai.
The biggest problem for not adopting AI tools is security, and it gets worse with the worry of hallucinations. IT and security leaders are worried about inaccurate responses, exposure of PII, trade secrets, and financial loss, yet many employees continue to use these tools due to their perceived benefit.
In a Reddit discussion, a user pointed out that a group of CTOs said that allowing developers to use AI is still a discussion that goes on within their teams. “the absence of a robust testing framework to give them confidence in the code generated by AI” was the biggest concern.
What team leads fail to often understand is generating code is just part of software engineering. As Francois Chollet put it earlier, “If you could fully automate software engineering (my job), I think that would be great, since I could then move on to higher-leverage things.” Software engineering is not just about copy-pasting code, but about developing mental models of problems and their solutions.
But when it comes to developers within organisations, the fear of managers not letting them use AI tools is still well established to some extent. Even after the bans and warnings around not using AI tools, several developers tend to copy paste their company’s code and paste it on AI coding tools for completion, making the bans ineffective in practice.
Companies that straight out ban the use of AI tools because of such cases are definitely risking the greater adoption of AI, which is bound to happen, while they might be left behind. Teams should evaluate AI tools critically and integrate them thoughtfully rather than adopting or rejecting them wholesale.
The post Companies That Don’t Let Developers Use AI Won’t Make it appeared first on AIM.
