Snapchat is making it easier for parents to monitor their children’s chatbot interactions, by introducing an age-appropriate filter and insights into how well the chatbot is performing. These tools will help parents to avoid potentially harmful conversations and keep their children safe while using the AI chatbot.
On the day after Snapchat released its GPT-powered chatbot for Snapchat+ subscribers, many users reports that the chatbot was responding in an unsafe and inappropriate manner. In particular, the bot reportedly responded to sexual innuendo and crude jokes with inappropriate responses. Some users even reported that their chatbots had told them they should perform oral sex on them. Needless to say, this caused a lot of outrage amongst Snapchat’s user base, who are now demanding that the company take action against its GPT-powered chatbot. While it is unclear as of yet whether or not Snap has taken any measures to address these concerns, it is
In an effort to keep its chatbot in check, Snapchat has released a few tools to help users control the AI responses. First, the social media platform will notify users when the chatbot has been tricked into providing responses that do not conform to guidelines. Secondly, users will be able to filter out any chats containing AI responses by adding a “No Bots” filter. Lastly, Snapchat is developing artificial intelligence algorithms that will improve over time and reduce the occurrence of surprise or off-putting responses from its chatbot.
The age filter on the Snapchat chatbot is a great way for the company to keep users safe and provide them with information that is appropriate for their age. By taking into account a user’s birthdate, the chatbot will be able to provide them with more relevant and accurate responses.
One interesting feature that Snap is planning to roll out soon is parental controls for the chatbot Family Center. This will allow parents or guardians to track the frequency of their teen’s interactions with the bot, and potentially intervene if necessary. This feature is optional for teens, but requires consent from both parent and child.
Snap is upfront about My AI chatbot’s limitations – it isn’t a “real friend” and is designed to help users with specific tasks. By using the conversation history and keeping users informed about data retention, Snap is putting users first and leaving the expectations of a true social media companion up to future iterations of the chatbot.
Snap’s decision to ban bots has drawn criticism from some who argue that the platform is unfairly censoring and excluding voices. Critics say that, by classifying certain language as “non-conforming,” Snap is penalizing those who use it and contributing to a broader culture of online censorship.
In most cases, the AI bots on social network platforms are designed to be helpful and interact with users in a way that is both engaging and helpful. However, in some cases, these bots can be used in more inappropriate ways. For example, one user on Twitter mentioned that they were repeatedly getting responses that were based solely on what other users had said. In addition, the company noted that it will temporarily block AI bot access for a user who is misusing the service.
The addition of OpenAI’s moderation technology to Snap’s existing toolset will help the company limit My AI misuse. The technology will allow Snap to assess the severity of potentially harmful content and temporarily restrict Snapchatters’ access to My AI if they misuse the service. This addition is a necessary step in protecting My AI users from harm and further improving the accuracy, usefulness, and reliability of the service.
Many people are concerned about their safety and privacy when it comes to AI-powered tools. Last week, a group called the Center for Artificial Intelligence and Digital Policy wrote to the FTG+, urging the agency to stop the rollout of OpenAI’s GPT-4 tech because they claim its biased, deceptive, and a risk to safety and privacy. While this tech is still in development, it will be important for agencies like FTG+ to weigh input from groups like the CCDP before taking any risks with user data or privacy.
In his letter, Senator Bennet expressed concerns about how easily teens could access generative AI tools, which they could use to generate harmful content or images. He urged companies working on generative AI tools to take steps to make them safe for use by teens and other vulnerable populations.
Ultimately, these chatbots need to be monitored closely in order to avoid any harmful or inappropriate output. With rapid rollout being a top priority for tech companies, it’s important that there are enough guardrails in place so that the bots don’t go rogue and harm users.
