Co-piloting the Olympics: “human-centric AI” as a creative partner

The Olympics has long embraced technological innovation that enhances the viewing experience and improves content creation. While concerns have emerged about how AI might disrupt industries, Olympic Channel Services (OCS) moved to leverage the technology, creating a human-centric Editorial Co-Pilot, designed to enhance human creativity rather than replace it.

The tool was created to address the specific needs of content creators covering a massive, fast-paced event like the Olympics. Unlike many AI applications that seek to automate entire processes, the Co-Pilot focuses on supporting human judgment and creativity.

“We guide [the editorial team] through the whole writing process, from brainstorming to the publishing of the content they have created,” explains Iñigo Cristóbal Losada, Artificial Intelligence Lead at OCS.

The AI team was clear that the Editorial Co-Pilot should be a tool to assist, not replace, human content creators. Losada explains that the objective was to ensure that AI would free up time for users, enabling them to focus on more creative tasks. “This is human-centric AI. The person is always in control of the content,” he says. Rather than generating content, the Editorial Co-Pilot makes suggestions, provides support, and simplifies workflows while leaving the final decisions in human hands.

This vision was particularly important during the Paris 2024 Games, when content creators were working under immense pressure to generate stories in real-time. Losada and his team recognised the strain this puts on editorial and social media teams, and their AI solution was designed to ease these pressures without removing the human element. “The key is that the person gets the suggestion, and if it works, great, if it doesn’t work, that’s fine,” he says.

Read more Paris 2024: OBS gets up close and personal to capture Olympic spirit

Editorial Co-Pilot is a Chrome extension integrated into the existing workflows of editorial teams. It guides content creators through three key phases of content creation: brainstorming, writing, and publishing. The tool pulls data from platforms like Semrush and Google News to provide relevant SEO terms, trending topics, and related questions during the brainstorming phase. It also checks to see if a story has already been covered. However, even in this early stage, the human creator remains in control, manually selecting from the tool’s suggestions.

Once an idea advances from the brainstorming phase, the tool can be used to suggest social media captions, check alignment with the style guide, list related articles, and suggest SEO terms, among other functions.

AI designed to enhance creativity

The guiding principle of the Editorial Co-Pilot is its ability to suggest, but not impose, ideas and tags for articles. When content creators are drafting stories, the tool can propose SEO-optimised keywords and even social media posts based on the article’s content.

For example, when writing an article, a content creator can highlight a section of text, and the Editorial Co-Pilot will suggest relevant tags for that content. This can be an uphill task in large-scale events like the Olympics, where hundreds of pieces of content are generated daily. The Editorial Co-Pilot automates this tagging process, reducing human error and saving time. Losada emphasises that users can edit or reject the tags if they feel they aren’t suitable. “The tool automatically gives you tags for athletes, or disciplines, or analysis,” he says. “But it’s also explaining why it is giving those suggestions. There’s context for them, so trust is built in the tool.”

Watch now AI in Action: Building a solid data foundation to enable AI

The same goes for social media posts, which the tool can generate for platforms like Instagram and X. During the Olympics, teams often need to publish dozens of posts in rapid succession. The Editorial Co-Pilot can draft these posts, but it’s up to the content creators to review, tweak, or discard them. This human/AI collaboration ensures that while repetitive tasks are automated, the nuanced, creative decisions remain firmly in human hands.

Enhancing workflows, not replacing human effort

Losada acknowledges that introducing any new technology to an established workflow -especially during an event as complex as the Olympics - can be daunting. This was one of the main challenges the team faced launching Editorial Co-Pilot only weeks before the Paris 2024 Games. “It’s hard to introduce new tools when you are getting close to something that you’ve been working on for months,” he says. “With the Olympic Games, because there are so many stories, the capacity to absorb new tools is limited. So, it was a bit of a challenge, and the approach was to constantly talk to them and build some trust, and at the same time, some hype in terms of what it could do.”

Despite these challenges, the tool quickly proved its worth. During the Games, the Editorial Co-Pilot helped streamline workflows in several ways, from suggesting push notifications to generating social media posts for medal wins. One feature that received positive feedback was the automatic generation of ‘medal posts’, which the tool populated with the athlete’s name, event, and result, allowing social media teams to publish celebratory content quickly. These pre-structured posts helped teams stay on top of the fast-paced nature of the Games, while still leaving room for human creativity and input.

Reception and building trust in AI

The reaction to the Editorial Co-Pilot from its users was generally positive and by building trust and being transparent about how the tool works, the team was able to gain user acceptance. By the time of the Paris 2024 Games, the service had around 42 weekly users, which included editorial, marketing, and social media teams. While this number wasn’t huge, Cristobal pointed out that it was a solid start, considering the size of the editorial team and the timing of the tool’s introduction.

Users appreciated the time-saving aspects of the tool, particularly when it came to automating repetitive tasks like generating tags and creating social media posts. The marketing team in particular provided positive feedback, noting that the tool saved them “hundreds of hours of content creation” during the Games. The Editorial Co-Pilot was especially useful for handling last-minute stories and unanticipated events, where quick content generation was essential.

Losada shared that the tool was most beneficial in situations where teams had little time, and the AI could step in to offer content suggestions, push notifications, or social media posts. In these high-pressure moments, the Editorial Co-Pilot provided valuable assistance without taking away the final editorial control from the human creators.

Future plans: iteration and human-led innovation

Losada emphasises that future development will be centred on continuous improvement and adaptation based on user feedback. He explains that the tool is “alive” and will keep evolving to meet the needs of content creators. The team’s approach involves regular updates and iterative changes, with a strong focus on enhancing the user experience.

During the Paris 2024 Olympics, the service introduced a feature called ‘interesting moments’, which analysed website traffic spikes to suggest relevant stories. This feature was tailored for the Games and has since been removed, but its development exemplifies how the tool can adapt to specific needs during different events. Losada stresses that the Editorial Co-Pilot’s functionality will continue to evolve as they gather more user feedback from editorial, marketing, and social media teams, ensuring it remains a valuable asset for all.

Looking ahead, Losada says that they plan to integrate the tool more deeply with social media platforms and enhance the brainstorming phase. These future developments will aim to further streamline workflows and improve collaboration between AI and human creators. The key to these updates is keeping the tool user-focused, ensuring it supports rather than replaces human decision-making, and refining it based on real-world usage.

Ultimately, he sees the Editorial Co-Pilot as a long-term solution that will continue to develop over time, with new features and improvements driven by the evolving needs of content teams.

AI as a creative partner

The development and success of the Editorial Co-Pilot at the Paris 2024 Games highlight a broader trend in AI technology - one that places humans at the centre of the process. Rather than viewing AI as a replacement for human work, the Olympics have used it as a collaborative partner. This human-centric approach ensures that while AI can enhance workflows and boost efficiency, the final product remains under human control.

As AI continues to evolve, the Olympics editorial team’s work with the Editorial Co-Pilot offers a blueprint for how other organisations might adopt similar tools. By focusing on enhancing, rather than replacing, human creativity, the tool demonstrates a balanced approach to AI implementation that will likely become more common in content creation workflows across industries. “I think that Paris was a huge milestone and will enable us to keep building more new things. We’ve built trust with our stakeholders, we’ve shown the things that can be done, that they work, and that they can be very useful.” Losada says.

Watch now Paris 2024 Olympics a hotbed for broadcast innovation