The Trust Algorithm

Driving a Future of Trust

Ford strongly emphasizes ethical AI use as part of its mission to help build a better world where every person is free to move and pursue their dreams.

Our AI Ethics Principles are based on 7 key pillars: Trust, Social Responsibility and Mobility; Safety; Fairness and Inclusivity; Transparency, Accountability and Explainability; Protection of Privacy; Oversight; and Best Practice. These principles provide a practical guide for Ford's employees on how to use AI responsibly and in compliance with existing laws, as well as in accordance with international best practices.

In practical terms this means we keep human-centered design and thinking in all of our AI use. We employ rigorous testing, validation, and auditing processes on an ongoing basis for all our AI technology — including auditing for bias and discrimination. And privacy-by-design is embedded throughout the entire AI product development cycle. As a thought leader, we seek to contribute to the development and definition of best practices and share our experiences with others, fostering collaboration and growth in the field of AI.

Racing for the Truth

In July 2023, U.S. President Joe Biden announced that 15 leading AI developers — including Alphabet and Microsoft — voluntarily committed to making AI content safer from fraudulent use. And on Oct. 30, 2023, Biden signed an executive order directing companies to uphold new standards for AI safety that include “protect[ing] Americans from AI-enabled fraud.”

One of the protective measures companies often discuss, and was specifically noted in the Executive Order, is watermarking AI-generated images. Watermarking adds invisible data to an image to mark it as having been created by AI. If someone tries to pass off an AI-made picture as real, the watermark will expose the fraud. In October, Microsoft integrated an AI-powered text-to-image creator in its Bing Image Creator and Chat services that watermarks each image with the date and time it was made.

However, a research team at the University of Maryland showed that it's possible to both remove watermarks (so that AI-created images won't be detected), and to add false watermarks to real images. Study authors Mehrdad Saberi and Soheil Feizi explained to the enterprise technology news publication, The Register, “We believe image watermarking will be a race between defenses and attacks in the future.”

Your AI Travel Partner

Personal AI travel assistant GuideGeek provides bespoke travel recommendations via conversational chats with its AI-powered bot in WhatsApp and Facebook Messenger. By sourcing a wide array of data sources that encompass hotel and restaurant reviews, and collaborating with Skyscanner for flight deals, GuideGeek's recommendations achieve an impressive 95% accuracy rate — meaning travelers can reasonably trust the app to steer them straight.

What's most notable about this AI travel buddy is that it dives into a traveler's preferences to create highly customized itineraries. GuideGeek's approach to travel planning is to augment the traveler's journey, not replace the human element that adds depth and authenticity to the travel experience.

With 1.3 million followers on Instagram, GuideGeek has some serious devotees — including travel influencer Michael Motamedi who traveled the world with his family for 12 months, letting GuideGeek's AI make the decisions every step of the way. Motamedi loved the experience and said that with GuideGeek his family was often “discovering unique things about a city in real time” that they would not have found on their own.

Swiping Right

Tinder is aiming to make its dating app safer by enhancing its Photo Verification feature. Tinder users will now be required to provide video selfies and complete a series of video prompts, instead of static photos, to confirm their identity. AI will analyze the videos to ensure they match their profile photos. This video-based liveness check is coming at the perfect moment to help combat the increasing ease of creating fake photos using generative AI.

The update will also grant users the option to restrict their chats exclusively to those who are verified by the AI tech, and for users willing to pay, they can choose to only see verified profiles. The feature will replace the previous option of uploading photos, and existing Photo Verified users will need to upgrade to the video selfie feature to maintain their verification status.