In October 2022, Bumble announced that its AI tool for detecting adult images will be available to other developers on GitHub. The aptly named Private Detector was rolled out in June 2019 to prevent unsolicited nudes from being shared through the chat function in the Bumble app, a practice known as ‘cyberflashing.’ By sharing this safety-enhancing AI tool with others, the internet can become a safer place for all that use it.

Crimes IRL should be crimes online, too

Just as it is illegal for individuals to expose themselves in public, it should also be illegal for individuals to expose themselves online without a prior invitation to do so. Consent is key. Cyberflashing is considered a crime in California and Texas (the latter after founder Bumble Whitney Wolfe Herd worked with legislators to get this changed), though not elsewhere in the US. In March 2022, cyberflashing became a criminal offense in England and Wales as part of the UK’s Online Safety Bill. The act was criminalized in Singapore in 2019 and is also illegal in France and Ireland. There is more work to be done on criminalizing this elsewhere.

Though this form of harassment is severely underreported, 2017 YouGov research found that 41% of millennial women had been sent an unsolicited image online, with cases including on public transport, often over Apple’s AirDrop. 2020 research by Professor Jessica Ringrose from the UCL Institute of Education found that 76% of girls aged 12 to 18 have been sent unsolicited nude images of boys or men. Though fault clearly lies with those sending unwanted images, AI has the potential to help protect people online.

Bumble training dataset

Bumble’s Private Detector automatically blurs potential nude images shared through the chat function on the Bumble app, reportedly with 98% accuracy. Recipients can then decide whether they want to view or block the image, and reporting unsolicited nudes is easy in the app.

The training data and algorithm are now available in Bumble’s GitHub repository. Though it is not the most sophisticated computer vision algorithm, it does not need to be complex to be an incredibly useful tool. Although a very minor proportion of Bumble’s users send unsolicited lewd images, the volume of data that Bumble has is enough to create a rich training dataset, which is critical for machine learning model training. The safety tool can be adapted and embedded into other social media apps and websites.

Open-source collaborations will spearhead AI development

Open sourcing algorithms helps to propel AI development and use. Although algorithms can be costly to design, a large hurdle for AI development is often collecting and labeling sufficient amounts of high-quality data. This is where Big Tech companies often have an advantage over small start-ups and academic institutions—they have the data. The center of AI development is shifting away from academia and towards companies that have the advantage of owning huge datasets.

How well do you really know your competitors?

Access the most comprehensive Company Profiles on the market, powered by GlobalData. Save hours of research. Gain competitive edge.

Company Profile – free sample

Thank you!

Your download email will arrive shortly

Not ready to buy yet? Download a free sample

We are confident about the unique quality of our Company Profiles. However, we want you to make the most beneficial decision for your business, so we offer a free sample that you can download by submitting the below form

By GlobalData
Visit our Privacy Policy for more information about our services, how we may use, process and share your personal data, including information of your rights in respect of your personal data and how you can unsubscribe from future marketing communications. Our services are intended for corporate subscribers and you warrant that the email address submitted is your corporate email address.

This is a trend that will likely continue in the years to come as the volumes of data generated in our everyday interactions only continue to grow. Open-source collaborations with institutions will pave the way for AI progress. The merits for start-ups may come from creating new case applications for open-source available algorithms and data sets.

Bumble shows tech needs diverse leadership

In February 2020, 31-year-old Wolfe Herd became the youngest female CEO and founder to take a US company public. This female-driven business joins a very short list—of the 559 companies that went public in the US in 2020, only two were founded by women. Bumble, a company built by women, for women, unsurprisingly has “women at the center”, as the IPO filing states.

Racism and misogyny are rife online. The internet is used by a diverse cross-section of the population, and it follows that those building it should also be diverse, representing the rich and wide range of people in society. Not only does diversity of leadership promote diversity of thought and helps generate new ideas, but it is also critical that those using technology have the internet built for them in a way that keeps them safe.