[ad_1]
The following is an excerpt from the 2022 Cisco Purpose Report, printed on December 8, 2022.
As the stewards of the info that make trendy life doable, expertise firms should earn prospects’ and customers’ belief. This means being aware of how our merchandise are made and used and taking steps to deal with potential damaging impacts. Cisco strives to design and construct expertise in ways in which respect human rights, promote inclusion, and defend privateness and safety—so that everybody can profit from a extra related world.
When it involves Powering an Inclusive Future for All, respect for human rights is a elementary innovation precept. Privacy, safety, and inclusion should be central within the design methodology, particularly in terms of synthetic intelligence and machine studying. Training knowledge units usually decide how product design and consumer expertise take form—the adage “goodness in, goodness out” actually applies. Developers should take steps to make sure these knowledge units are sturdy, numerous, and consultant of all customers. Failing to take action may end up in inaccuracies, disappointing consumer experiences, and unintended bias.
Take, for instance, Webex digital backgrounds, that are designed to cover customers’ environment to reinforce privateness, safety, professionalism, and enjoyable. Early analysis variations of this function, which mirrored the state-of-the-art on the time, didn’t carry out properly for sure hair textures and types, or lighting situations. In some instances, they inadvertently filtered out parts of a consumer’s look. Our engineers acknowledged that broader, extra numerous coaching knowledge units that had been consultant of the Webex consumer base had been required, and addressed this through the design section earlier than releasing the function. By making use of knowledge that was anonymized, ethically sourced, user-contributed, open-source, supplied with express consent, and respectful of particular person privateness in different methods, we made the coaching knowledge extra sturdy. This resulted in additional consultant photos and algorithms, and a significantly better and extra inclusive consumer expertise for all.
In 2022, we constructed on the learnings from the Webex group and launched our Responsible AI Framework primarily based on six rules: Transparency, Fairness, Accountability, Privacy, Security, and Reliability. Our Responsible AI Working Group constantly drives adherence to those rules by placing new applied sciences by way of Responsible AI Impact Assessments, providing steering on the right way to handle danger to human rights, and offering accountability by way of incident reporting of human rights, privateness, and safety considerations.
To be taught extra in regards to the progress we’re making to energy a extra inclusive future, go to our Cisco ESG Reporting Hub, the place you possibly can learn our 2022 Cisco Purpose Report.
Share: