Microsoft has launched a preview of its Copilot Vision feature, allowing users to enhance their browsing experience through AI assistance. The feature is currently available to Copilot Pro subscribers in the U.S., priced at $20 per month. Copilot Vision enables the AI to observe web pages users visit, facilitating interactive support while browsing. Microsoft emphasizes a strong commitment to privacy, ensuring that data from user sessions is deleted immediately after use.
“Since we first announced the new consumer Copilot in October, we have loved seeing the ways people have made Copilot their own. It has helped people think through their ideas and get a fresh perspective, has made it less overwhelming to find the right information, and has even helped people feel supported when navigating difficult conversations.”
-Microsoft
Microsoft launches Copilot Vision for AI-assisted browsing
Copilot Vision integrates into Microsoft Edge, providing real-time assistance as users navigate various sites. Designed for an opt-in experience, it allows users to engage with the AI to discuss content, summarize information, or answer questions related to their online activities. Microsoft described the feature, stating, “When you choose to enable Copilot Vision, it sees the page you’re on, it reads along with you, and you can talk through the problem you’re facing together.”
The tool’s capabilities extend beyond simple query responses; it can summarize text, translate languages, and assist in various tasks, including identifying discounted products or offering gaming strategies on platforms such as Chess.com. However, the initial rollout of Copilot Vision is limited to a specific selection of websites, excluding paywalled and sensitive content. Microsoft clarified that the allowed websites will be determined on a case-by-case basis, with an intention to expand the list over time.
As part of the ongoing launch, Microsoft is taking precautions to avoid past privacy issues related to AI tools. The company assures that Copilot Vision will not save or utilize processed audio, images, or text for training its AI models. This is a strategic move to alleviate concerns among users and publishers regarding data privacy and usage.
Investigations into the implications of Copilot Vision’s capabilities are ongoing, especially given the legal challenges technology companies currently face in regard to data usage on news and media sites. Microsoft faces scrutiny from major publishers, like The New York Times, regarding the potential infringement of paywall protections through the use of AI tools to access their content. Microsoft has indicated that it will respect “machine-readable controls” designed to limit unauthorized AI access, although specifics on these controls remain unclear.
Furthermore, the company’s cautious approach reflects its desire to foster trust among publishers and users alike. Microsoft stated that feedback from various sources is crucial for developing Copilot Vision in a manner that is both user-friendly and compliant with existing web policies.
The feature’s launch is a tactical measure by Microsoft to position its Copilot offerings more effectively in the competitive landscape of AI browsing tools. As Copilot Vision continues its limited deployment, Microsoft will likely monitor user interactions and publisher responses closely, adjusting its approach as needed.
With privacy concerns at the forefront of technology discourse, the reception of Copilot Vision may significantly influence future AI features within web applications. Microsoft aims to maintain a balance between innovation and ethical considerations as it navigates the complex terrain of AI capabilities.
Continuing developments will shed light on Copilot Vision’s impact on browsing behavior and its interaction with existing web standards, keeping the tech community attentive to both user experiences and industry regulations.
Featured image credit: Microsoft
