MSN, recently known as Microsoft Start, is a news and content platform that provides users with news articles, entertainment, lifestyle content, and more.

organization

Microsoft AI  

(Windows, Bing, Copilot, MSN)

role

Design, Development,

Research

collaborators

Developers, Project managers,

UX researcher

timeline

2 months

overview

Designing for Trust and Engagement in an AI-Powered Commenting Experience

I collaborated with product and engineering to redesign the commenting experience on MSN, aiming to boost participation and build trust around AI-generated content (AIGC). Our goal was to make AIGC feel more reliable and engaging by rethinking how information, polls, and discussions surfaced, through clearer signals, more visible activity, and lower-friction entry points. This included new features like contexual AI generated poll questions, AI generated summaries and quick comments.

The challenge

Users are reading news, but skipping the conversation

The challenge was to design a space where users wouldn’t just passively consume new stories but feel confident and motivated to share their own perspectives.

This meant rethinking the way information was structured, how polls and discussions were surfaced, and how the system could better invite engagement such as a pop-up card, clearer content signals, more visible community activity, or frictionless comment entry points.

Pain Points and Business Goals:

Passive consumption of content without engaging with the comment panel

Skepticism toward AI-generated content and insights

Low participation in user discussions

problem framing

Understanding the engagement gap

Even though there was strong engagement with polls, we noticed a significant drop-off when it came to users commenting or continuing the conversation. Users were reading AI-generated market summaries but not contributing their own insights.

Through internal feedback and behavioral data, we identified a few key friction points:

Lack of trust in AI content: Users weren’t clear what role AI was playing in the summary or how accurate it was.

Unclear calls to action: The transition from reading to commenting lacked a clear bridge.

Thin sense of community: Without visible signals of ongoing conversations, the page felt static.

Opportunity: How might we redesign the discussion and pop-up card experience to feel more interactive, credible, and participatory, especially in a space where users may feel uncertain about AI-generated content?

My role

My impact

As the UX designer, I led the end-to-end re-design across four core initiatives:

Discussion Page and Pop-up Card Enhancements:
Introducing a clearer hierarchy with AI insights, poll data, and discussion stats to drive transparency and encourage exploration.

AI-Generated Comment Suggestions:
Integrated GPT-powered comment prompts aligned with user poll answer to reduce friction on commenting. Built safety mechanisms for human review to ensure trustworthy, bias-mitigated suggestions.

Design and Research Study on AI Labeling:
Designed and conducted a study on label placement and wording to improve user understanding and trust in AI-generated content. Insights directly informed key design decisions.

Information Density & Discoverability:
Balanced readability and depth by truncating and expanding content with contextual interactions.

Solution: canvas refinements

Canvas Refinements

Through rapid iteration and user research, I transformed initial product hypotheses into UX solutions that increased engagement, strengthened community interactions, and promoted transparency around AI-driven content.

Before

After

Establishing AI Credibility Through Citation Placement

While research clarified user understanding of AI-generated content, it raised a new question: how do we visually signal AI authorship in a way that feels both intuitive and trustworthy?

To align AI-generated insights with established patterns of credibility, I advocated for placing the AI label in the top right corner of the poll module, mirroring academic citation formats like APA and MLA. These standards consistently display source information in the top right, helping users recognize and trust the origin of content quickly.

APA Journal Article Format
Journal Title + Year, Volume, Issue, Page Range
Top right corner

MLA Format
Author's Last Name + Page Number Shown
Top right corner

My Rationale

By borrowing visual cues from familiar scholarly formats, we could help users more intuitively understand that the poll summary was AI-generated, reinforcing transparency and perceived legitimacy.

However, after discussion with the team we decided for a left-aligned label, citing reading patterns and F-shaped scanning behavior (users tend to read from left to right). This sparked a valuable debate about whether to prioritize visual familiarity (top-right = source) or behavioral readability (left-first scanning).

Why It Mattered

After discussing my proposal with the product team, we ultimately placed the label on the left for better scannability. The discussion sharpened our shared understanding of information hierarchy and trust signaling.

This led to further design iterations and user testing, including testing variations in placement and label phrasing, to ensure users could clearly and confidently identify AI-generated content without interrupting their reading flow.

Research and insights

Before locking in design decisions, I realized we needed to validate how users understood AI involvement. That led to a targeted UX lab focused on label clarity. Through rapid iteration and user research, I transformed initial product hypothesis into UX solutions that increased engagement, strengthened community interactions, and promoted transparency around AI-driven content.

Hypothesis: Consistent and well-placed labels would build clarity and trust.

To test my hypothesis around AI transparency, I designed and conducted a lab study to explore how labeling affects user understanding of AI-generated content (AIGC).

I tested several iterations of designs to inform key design decisions and formulated targeted questions to probe whether placement and copy of the labels influence users’ ability to distinguish between human and machine-written content.

The core research question:

Does the placement and copy of the labeling affect users’ understanding of what is AI-generated content (AIGC)?

Findings from this study informed key design decisions around transparency, trust, and responsible AI integration.

Scenario:
Imagine that you have answered a poll question about the U.S stock market.

Questions:
Parts of this experience are powered by AI. Which do you think are powered by AI? Take a look at the image and list any parts that you think are powered by AI.
Why?

degradowski

“I think it is possible the Background section was written by AI. It would make sense to have AI do this to save time. I also assume that the system for determining which comments are Top Responses is handled by AI in order to recommend the best comments.”

laurynworley

"I would say that the 'Background' part definitely came from AI, since the top right says 'Insights from AI' and it is the only section that can be considered 'insights'."

warris

"The poll itself and the 'background' section could have been generated by AI."

Results = 36 total participants

Poll question = 14%

Background = 56%

Both = 31%

🔍 Key Insights from 36 Participants:

Most users recognized the “Background” summary as AI-generated (56%), while fewer identified the poll question alone (14%). Notably, 31% believed both were AI-driven, indicating that once AI is mentioned, people tend to assume deeper involvement. The “Insights from AI” label heavily influenced perception, even if the content wasn’t AI-made.

Majority Identified the Background Section as AI-Generated

56% of participants believed the "Background" section was AI-generated.

Reasons included:

It saved time and appeared to be auto-generated.
It aligned with the “Insights from AI” label at the top.
It matched expectations of summarization tasks suited for AI.

Some users assume AI powers the entire experience.

31% believed both the poll question and the background were AI-driven, which suggests a general openness to AI involvement across different UI elements when insights are presented.

Few attributed AI to the poll alone

Only 14% thought the poll question itself was generated by AI, reflecting a perception that polls feel more user-initiated.

AI in Comment Curation

Some participants speculated AI was also involved in ranking or recommending top comments, indicating an understanding of how AI may be embedded beyond visible content.

Design Implications

Users recognize AI-generated content when it fits familiar patterns like summaries or statistics. Labels such as “Insights from AI” strongly shape perception, even if the content isn’t AI-made. Clearer signals separating human and AI input can build trust, especially as users increasingly expect AI to drive both content and curation.

Why It Matters

Users were more likely to trust and engage with the content once they understood what was created by AI. Clarifying attribution helped reduce skepticism, reinforce transparency, and create a foundation of trust that’s crucial in news contexts involving AI summarization.

UX Labs 2: Testing Label Clarity: 'Insights from AI' vs. 'Powered by AI'

To understand how users interpret AI labels, I ran a targeted A/B test comparing two framing approaches. Initial feedback on our metadata labels revealed friction. To unpack this further, I ran a study comparing three UI treatments to understand which one best supported scannability and comprehension.

To understand how users interpret AI attribution labels, we tested two design variations:

Concept A: “Insights from AI” (text label)

Concept B: “Powered by AI” with a sparkle icon

Scenario:
Imagine you've answered a poll relating to your thoughts on the U.S stock market. AI generates both the poll and the background information.

Question:
Which of these images best show that?

Concept A: Insights from AI Top Right

Concept B: Powered by AI Sparkle Icon Top Right

🔍 Key Insights from 43 Participants (p=0.033)

While Concept B ("Powered by AI") was slightly preferred, the result was not statistically significant (p=0.033). However, feedback from participants offered valuable insight into how language and visual framing shape trust and understanding.

While both labels communicated AI involvement, participants saw “Powered by AI” (with a sparkle icon) as clearer, more system-driven, and trustworthy. “Insights from AI” felt editorial and vague to many, which created confusion about what the AI actually contributed.

Design Decision

I adopted “Powered by AI” as the preferred label and paired it with a recognizable sparkle icon. You also improved visual hierarchy to make the labeling more scannable and cohesive with the rest of the UI.

Even without a conclusive winner, this test helped surface users’ mental models of how AI works in the experience, and it informed our next iteration on AI transparency and labeling.

Why It Matters

Stronger and more explicit AI attribution improved users' mental models of how the system worked, reinforcing transparency. This helped users feel more confident in what they were reading—and made AI involvement feel intentional rather than hidden.

UX Labs 3: Clarifying Poll Metadata: Text, Icons, or Both?

After refining how we labeled AI-generated content, the next challenge was ensuring users could clearly interpret poll-related interaction metadata, like how many people voted, commented, and when the poll was posted.

While the trust label helped clarify AI's involvement, users still struggled with understanding basic engagement signals, especially when labels were missing or icons were unclear. To address this, I conducted a third UX lab to test different treatments for poll metadata legibility and scannability.

Scenario:
Imagine you've answered a poll question about the U.S stock market and you want to know how many total people voted and commented, and when the poll was posted (shown under the poll question).

Question:
Which treatment do you think best shows this information?

🔍 Key Insights from 150 Participants (p=0.003)

With 150 participants, Concept A emerged as the statistically significant winner (p=0.003), beating both B and C designs.

Easy Scan and Understand

Users strongly preferred Concept A (icon + text), saying it made poll activity (votes, comments, time) more scannable and understandable at a glance.

lonpatriot

"Concept A is my preferred choice because I can quickly read and analyze the poll and see the total number of votes... Concept B would leave me slightly confused and would take me longer to interpret."

Icons Alone Create Ambiguity
Concept B (icon only) led to confusion. Users didn’t always recognize what the symbols meant without accompanying labels.

gwynpatadia

"While the bar and dialogue symbol makes sense, I feel like it could be a little bit more confusing for newer users."

Text-Only Requires Extra Effort

Concept C (units only) was readable but felt dated and harder to parse quickly. Users missed the visual cues that help with scanning and recognition.

laurynworley

"Concept C is too normal as it only utilizes text. This is an old way of presenting the same information."

Words Like "Votes" and "Comments" Are Necessary
Microcopy mattered: participants explicitly mentioned needing unit labels (e.g. "votes") to grasp the meaning behind the numbers. Their absence made the experience feel incomplete.

barparde

"I prefer seeing 'votes' next to the number... it's especially helpful to see that 'votes' text to convey the message clearly."

lylaohare

"Concept A allows the users to identify the meaning of the numbers and the sections a little bit earlier."

Design Outcomes

Final Direction: Adopted Concept A — combining icons with unit labels for all metadata (votes, comments, timestamp).

Microcopy reinforced for clarity — retained explicit words like "votes" and "comments" instead of relying on interpretation.
Improved UI legibility and hierarchy — ensured that interaction metadata is both easy to spot and semantically clear, especially for quick-glance use cases.
Standardized treatment across surfaces — this pattern became the baseline for poll modules and future community interaction components.

Establishing AI Credibility Through Citation Placement

While research clarified user understanding of AI-generated content, it raised a new question: how do we visually signal AI authorship in a way that feels both intuitive and trustworthy?

I advocated for placing the AI label in the top right corner of the poll module, mirroring academic citation formats like APA and MLA To align AI-generated insights with established patterns of credibility, These standards consistently display source information in the top right, helping users recognize and trust the origin of content quickly.

APA Journal Article Format
Journal Title + Year, Volume, Issue, Page Range
Top right corner

MLA Format
Author's Last Name + Page Number Shown
Top right corner

My Rationale

Why It Mattered

While we ultimately placed the label on the left for better scannability, the discussion sharpened our shared understanding of information hierarchy and trust signaling. It also led to further design iterations, including testing variations in placement and label phrasing, to ensure users could clearly and confidently identify AI-generated content without interrupting their reading flow.

AI Suggested Comments

To complement the AI-generated summaries and polls, I explored how AI could also support users in participating more easily. One promising direction was using poll answers to generate suggested comments. These quick prompts lowers the barrier to entry for engagement, especially for users who didn’t know what to say but wanted to contribute.

Before

After

I raised concerns that letting users post AI-suggested comments without adding their own input could lead to low-effort responses and bot-like behavior. To address this, the PM and I aligned on a solution: use full sentence suggestions to kickstart empty threads, and shift to unfinished prompts in active ones to encourage more thoughtful, user-generated input.

Encouraging conversation With AI Suggestions when there are no comments

Thread State

AI Suggestion Type

Empty Thread

Full sentence prompts to spark engagement based off of poll answer

5+ Comments

Unfinished phrases to encourage user participation

Goal: Balanced ease of entry with authenticity
Impact: Increased trust, reduced bot-like behavior, improved comment quality

Encouraging conversation With AI Suggestions when there are no comments

Early in the design process, we explored giving users full, AI-generated comments to post with a single click. While this seemed like a fast way to boost engagement, I flagged concerns about authenticity, trust, and the potential for bot-like behavior dominating conversations. This frictionless approach risked turning discussions into a feed of AI-authored text, undermining our goal of building genuine dialogue.

To solve this, I proposed a nuanced alternative: dependent clauses generated from poll answers. Instead of complete sentences, we offered open-ended prompts that users had to finish themselves. This small design choice introduced just enough friction to make responses personal, without overwhelming users.

The impact was clear, threads maintained a human tone, and early usability feedback showed users felt “guided, but still in control.” This approach also aligned with our trust and quality metrics by reducing automated-looking comments and encouraging thoughtful participation.

Projected Engagement Lift: +12% in comment activity during early A/B tests.

Quality Signals: 28% more comments with original phrasing versus full-sentence AI suggestions.
User Feedback: “Helpful without feeling canned,” “Makes me think before posting.”