Listen to this story
|
OpenAI, ChatGPT’s parent, has launched a bug bounty program to incentivise researchers to detect and report any vulnerabilities in its artificial intelligence systems. The announcement was made through a blog post on Tuesday, revealing that the rewards would range from $200 for minor issues to $20,000 for “exceptional discoveries.”
“This initiative is an essential part of our commitment to developing safe and advanced AI,” wrote Matthew Knight, OpenAI’s head of security in the blog. “As we create technology and services that are secure, reliable, and trustworthy, we would like your help.” Moreover, the program does not cover instances of incorrect or malicious content produced by OpenAI’s technology.
The decision for the program does not come as a surprise. It was prompted by a recent incident involving Alex Albert, a 22-year-old jailbreak prompt enthusiast who managed to bypass the safeguards on ChatGPT.
OpenAI’s president and co-founder, Greg Brockman, in response, tweeted that the company had been considering a bounty program or network of “red-teamers” to identify weak spots in their AI models.
Democratized red teaming is one reason we deploy these models. Anticipating that over time the stakes will go up a *lot* over time, and having models that are robust to great adversarial pressure will be critical. Also considering starting a bounty program/network of red-teamers! https://t.co/9QfmXQi9iM
— Greg Brockman (@gdb) March 16, 2023
Sourced from the Open
Looks like the closed-door company is now desperately attempting to go open source route to improve its systems’ security posture with the recent bounty program announcement. However, it does not want to give back to the community.
For instance, its last open-sourced Gym and Universe seven years ago barely exists.
Another open-source initiative falls victim to the money conundrum is GPTx, where the company has been using open source to train and scale the models, it closed the doors on the code.
It is important to note that ChatGPT still uses open source code. Just last month when a security flaw exposed users’ search history, Sam Altman, OpenAI’s CEO acknowledged that the bug was caused by an issue in Redis client library.
we had a significant issue in ChatGPT due to a bug in an open source library, for which a fix has now been released and we have just finished validating.
— Sam Altman (@sama) March 22, 2023
a small percentage of users were able to see the titles of other users’ conversation history.
we feel awful about this.
So, like many others ChatGPT owes a great debt to the open source community. But the company seems to have turned a blind eye towards the community since ChatGPT is not open sourced, nor is it likely to ever be.
Moreover, the company’s previous product which caused a stir in the art community was DALLE.2 in Unfortunately, OpenAI is yet to launch an open-source version of DALL.E 2. The model remains exclusively available to a number of users who secured a spot on the waitlist.
OpenAI Misleads
Apart from its money making machine ChatGPT, the company has also been an important subject of discussion for being a closed source company contrary to its name, ‘Open’AI. In 2019, the former nonprofit restructured as a “capped-profit” and received a billion dollar investment from software giant, Microsoft.
Some worried that this move may result in making the innovative company no different from the other AI startups out there. A year later, Elon Musk, who was one of OpenAI’s co-founders tweeted, “I have no control & only very limited insight into OpenAI,” adding his confidence in its safety was “not high.”
Since he left the board in 2018 he has been actively vocal about his views on the company. The Twitter owner recently took to Twitter to express his concern over OpenAI’s shift towards becoming a closed-source, for-profit company controlled by Microsoft.
Musk, who had named OpenAI as an open-source, non-profit organization to counterbalance Google, criticized the move and urged the AI company to remain true to its original vision. “Not what I intended at all,” he said when Microsoft extended its long-term partnership with OpenAI through a new “multiyear, multibillion dollar investment.”
The bottom line is that even though OpenAI wants to remain closed door and not share their secret recipes, the company seems to not mind seeking help from the public to mend their ways.