KIMP’s Picks March 13th: Weekly Highlights In Design, Marketing, AI, & More

The AI realm is shifting rapidly. Blink and you miss a dozen updates that might affect your industry. Whether you are a marketer looking to keep up with the changes that AI is bringing into marketing or an entrepreneur trying to keep up with what’s changing or someone who likes to keep up with the social media buzz, we have good news. Staying in the loop is not that difficult anymore. Your weekly social media recap is ready. 

Welcome to the KIMP’s Picks’ March 13th edition – your shortcut to all the highlights. From big news in the AI landscape to social media algorithm changes coming your way, we have all the key updates right here. 

In the Marketing Realm 

Re-engaging old email lists 

MarTech shared a post discussing effective strategies to revive old email lists. As paid ads and organic social media are no longer enough, many companies are turning back to their CRM databases to revive old email lists. But emailing dormant contacts without a careful strategy can damage the brand’s reputation or worse, hurt their inbox placement. 

The post discusses how the biggest risk comes from outdated or inactive email addresses. Old lists often contain invalid emails, spam traps, or users who no longer recognize the brand. Email providers like Google closely monitor unusual spikes in sending volume. So, if a brand suddenly emails large inactive segments, it can also trigger spam filtering.

Taking all these into account, the post discusses tips like the proper configuration of DNS, identifying the right offers or messages for personalization of emails and others. So, if you are looking for ways to boost your email marketing strategies, this is a great read. 

What brands should know about AI rewriting digital marketing 

A recent post from Search Engine Journal discusses how AI is transforming the digital marketing realm. The marketing world is debating whether AI tools like ChatGPT will show ads. But the bigger shift is happening elsewhere as AI is quickly becoming the layer that helps users make purchase decisions online.

Also, platforms like Google, Microsoft, and Meta Platforms are embedding AI into search, shopping, and workflows. Instead of simply showing options, these systems increasingly compare products, shortlist recommendations, and guide users toward a final choice.

That changes how marketing works. For years, brands competed for visibility in search rankings or social feeds. But AI is changing that. Customers might ask an AI assistant for recommendations and these often become the search results page. 

This shift means brands must optimize not only for people but also for AI systems that increasingly influence buying decisions.

A new high-engagement advertising channel

Digiday shared a post recently exploring an insightful discussion regarding an emerging high-engagement advertising channel – in-flight advertising. 

With advertisers constantly battling ad fatigue and low recall, in addition to ad fraud online, there is a growing demand for more reliable advertising channels. One channel quietly standing out is in-flight advertising.

Ads delivered through airline entertainment portals and Wi-Fi platforms reach a captive audience. Additionally, passengers are seated for hours with limited distractions, which increases attention and recall. Unlike the open web, these environments are controlled, meaning every impression comes from a real person rather than bots.

In-flight audiences are also highly valuable. Airline passengers tend to be younger professionals, educated, urban, and more likely to fall into higher income brackets. Advertisers can also align campaigns with specific routes, events, and travel patterns. This signals a growing opportunity to reach engaged, premium audiences in a rare environment with guaranteed attention.

Ads on ChatGPT welcome a new demand-capture channel 

Talking about the quest for advertising channels with better engagement, there’s another one that brands need to think about and this post from Search Engine Land explores this. 

OpenAI has started testing ads inside ChatGPT with a small group of U.S. users. The sponsored placements are clearly labeled, but the move shows that there is a huge shift as AI platforms are exploring new revenue models.

If users begin shopping through AI prompts, ChatGPT could evolve into a new demand-capture channel similar to search. Instead of browsing multiple websites, a user might simply ask a question and receive a recommendation along with a purchase link.

But then this channel comes with its own challenges. Most AI queries today are informational rather than purchase-driven. And even when people research products through AI, they often complete the purchase on platforms like Amazon or Google, which makes attribution harder.

If adoption grows, ChatGPT advertising likely won’t expand the ad market. Instead, it will redistribute budgets currently spent on search and social.

Taking all this into account, the real opportunity lies in preparing for AI-driven shopping while testing the channel cautiously as it matures.

In the Content Marketing Realm 

LinkedIn content becomes a key source of AI answers 

If your content marketing strategy doesn’t prioritize LinkedIn yet, now is a good time to start. The post here from Semrush is sure to change your perspective. 

A new study by Semrush shows that LinkedIn is emerging as one of the most important sources shaping AI-generated answers.

Researchers analyzed 89,000 LinkedIn URLs cited by AI tools, including ChatGPT, Perplexity AI, and Google AI Mode. The data reveals LinkedIn is the second most cited domain across these platforms, appearing in about 11% of AI responses on average.

The study also found that AI models often mirror the meaning of LinkedIn content when generating responses. Also, original content performs best. Long-form LinkedIn articles between 500 and 2,000 words dominate AI citations, while mid-length posts of 50 to 299 words are also frequently referenced.

This shows that LinkedIn content is no longer just for social reach. It’s increasingly influencing how AI platforms explain and recommend businesses online.

In the Design Realm 

Air France’s new campaign uses design to redefine elegance 

Air France has launched a new global advertising campaign to highlight its ongoing shift toward a more premium travel experience. Visuals are a strong way to demonstrate a brand’s repositioning and Air France’s recent campaign highlights this.  

The campaign builds on the airline’s brand platform “Elegance Is a Journey”, introduced five years ago to support its move upmarket.

The new campaign showcases how French elegance and craftsmanship shape the airline’s travel experience. The campaign features 10 visual concepts and films designed to bring this promise of elegance to life. 

Each visual uses imaginative props created by French designers. Highlights include a seat shaped like the Wi-Fi symbol, a giant macaron celebrating French culinary expertise, and a quilt-style dress representing the comfort of Business class bedding. 

This campaign shows how storytelling and design can reinforce a premium positioning while showcasing product innovation

Canva introduces a new “Magic Layers” feature 

Canva has launched a new AI feature called Magic Layers that converts static images into fully editable designs. The feature is powered by the platform’s Canva Design Model and is currently available in public beta in the United States, the United Kingdom, Canada, and Australia.

Recently, AI image tools have made it easier to generate visuals quickly. However, most outputs remain flat files like JPEGs or PNGs. Once created, text and design elements are merged together, making edits difficult. Hence, even small changes often require generating a new image from scratch.

Magic Layers aims to solve that problem by reading a visual and rebuilding it as structured layers inside Canva. The system identifies elements such as text, objects, and backgrounds, turning them into editable components. Users can then adjust fonts, move objects, change colors, or refine layouts without starting over. 

In the AI Realm 

Perplexity announces “Personal Computer” 

After introducing Perplexity Computer a couple of weeks ago, Perplexity added a Voice Mode to it just last week. Now the company is expanding Perplexity Computer with the introduction of Personal Computer, a system that runs on a dedicated Mac mini and works around the clock as a digital proxy. 

Perplexity’s Personal Computer can access local files, connect with apps, and continue working even when the user is offline. Security features include approval prompts for sensitive actions, audit logs, and a manual kill switch.

Perplexity is also launching Computer for Enterprise, which connects AI directly to workplace tools like Snowflake, Salesforce, and HubSpot. Teams can query data, automate workflows, and generate dashboards or reports directly within platforms like Slack.

Google unveils Gemini Embedding 2 

Google has launched Gemini Embedding 2, its first fully multimodal embedding model. The model is designed to bridge the gap between different types of media. 

Traditionally, AI models struggled to process text, images, and audio in a single workflow. This new architecture maps everything from 120-second videos and PDFs to native audio and text in over 100 languages into a unified “semantic space”. 

The model can also understand inputs with diverse layers, such as an image paired with a descriptive text caption, to capture complex relationships. Early adopters like Paramount Skydance report significant improvements in media search accuracy when using the model.

OpenAI acquires Promptfoo

OpenAI is acquiring Promptfoo, an AI security platform that helps companies test and secure AI systems during development. The technology will be integrated into OpenAI’s enterprise platform, OpenAI Frontier, which is designed for building and managing AI coworkers.

As businesses continue to deploy AI agents into real workflows, security and governance have become critical. Enterprises need ways to test how AI behaves, identify vulnerabilities, and track risks before systems go live.

Promptfoo provides tools for evaluating and “red-teaming” AI applications. This helps developers detect issues such as prompt injections, jailbreak attempts, data leaks, and unsafe agent behavior. 

Mozilla x Anthropic partner to strengthen Firefox security 

Researchers at Anthropic, working with Mozilla, found that the AI model Claude Opus 4.6 can rapidly detect critical software vulnerabilities. In a two-week collaboration, the model identified 22 previously unknown bugs in the Mozilla Firefox codebase.

Mozilla classified 14 of those findings as high-severity issues, representing nearly one-fifth of all high-severity Firefox vulnerabilities fixed in 2025. Most of the issues have already been patched and released to users in Firefox 148.

The AI scanned thousands of lines of browser code and flagged potential weaknesses. This includes memory vulnerabilities that could allow attackers to manipulate data. 

The collaboration demonstrates how AI can accelerate vulnerability discovery while still requiring human oversight for validation and patching.

Anthropic unveils Code Review – a multi-agent system to detect bugs 

Anthropic is tackling the developer bottleneck with Code Review. This new multi-agent system is designed to catch the critical bugs that humans often overlook. Unlike traditional tools that prioritize speed, this system dispatches a team of AI agents to perform deep, parallel analysis on every Pull Request (PR).

The tool is currently available in research preview for Team and Enterprise users. The launch comes as AI-assisted coding rapidly increases developer output. At Anthropic, code production per engineer has grown by about 200% over the past year, turning code review into a major bottleneck. Code Review addresses this by deploying a team of AI agents to inspect each PR. 

Google expands Gemini in Chrome to more countries 

Google recently introduced a host of Gemini 3.1-powered features in Chrome to a limited market. And now these features are expanding to users in India, Canada, and New Zealand. The update also adds support for more than 50 languages, including Hindi, French, and Spanish.

Gemini in Chrome acts as a browsing assistant that works directly inside the browser. Users can chat with the AI without leaving their current tab, making it easier to execute tasks like summarizing articles, generating quizzes for study and more. 

The assistant also integrates with other Google services like Gmail, Google Maps, Google Calendar, and YouTube. This allows users to schedule meetings or even draft emails directly from the browser’s side panel.

Gemini in Google Workspace gets an upgrade 

Google Workspace is expanding its AI capabilities by turning Gemini into a collaborative partner across Google Docs, Google Sheets, Google Slides, and Google Drive.

For instance, in Docs, a new Help me create feature can generate a fully formatted first draft from a simple prompt. Gemini pulls context from emails, files, chats, and the web to structure documents instantly. 

Sheets now allows users to build or edit spreadsheets using natural language. Slides is also getting smarter. Gemini can design slides with clear messaging and balanced layouts while matching the style of the existing deck.

Drive is evolving from file storage to an AI-powered knowledge hub. Gemini can deliver quick answers, summaries, and insights from files, emails, and chats directly in search results.

Amazon’s Health AI agent provides virtual care 

Amazon has introduced a new AI-powered health assistant designed to simplify how people access and manage care. Called Health AI, the tool works within the Amazon One Medical app and the Amazon app to provide 24/7 health guidance. It can explain medical records, and also help users book appointments and manage prescriptions.

Built to provide users with virtual care, the assistant can answer health questions, interpret lab results, and offer personalized insights based on a user’s medical history (once they grant permission). Additionally, it can also connect users directly to healthcare providers through message, video, or in-person visits via Amazon One Medical.

Prime members also get five free direct-message care visits for common conditions like flu, allergies, and skin care. 

Google Maps gets a major AI-powered upgrade 

Google is rolling out a major AI-powered upgrade to Google Maps. This brings conversational search to Google Maps with the Ask Maps feature. This lets users ask real-world questions directly inside the app. 

Additionally, the update brings Immersive Navigation to Google Maps. According to Google, this is the biggest redesign of Maps navigation in more than a decade. The feature adds detailed 3D visuals showing buildings, terrain, and road structures. It highlights important driving details like lanes, crosswalks, traffic lights, and stop signs to help drivers make better decisions on the road.

Ask Maps is rolling out in the U.S. and India first, while Immersive Navigation is launching across the U.S. on mobile and in-car platforms.

Social Media Feature Updates and Algorithm Changes

Enhanced AI safety tools across Meta platforms 

Meta is introducing new anti-scam tools across Facebook, Messenger, and WhatsApp as part of a broader push to combat increasingly sophisticated online fraud.

The company is deploying advanced AI systems that analyze text, images, and contextual signals to detect scams such as celebrity impersonation, brand impersonation, and fake websites designed to mimic legitimate domains. These systems aim to identify deceptive activity earlier and at scale.

This also includes new user-facing alerts. Warnings for suspicious friend requests on Facebook, and suspicious device-linking requests on WhatsApp are a few to begin with. 

YouTube expands AI deepfake detection tools to protect public figures 

YouTube is expanding its AI likeness detection tool to help public figures identify and manage deepfake content on the platform. The feature, first launched for creators in the YouTube Partner Program, is now being piloted with government officials, journalists, and political candidates. 

The tool works similarly to YouTube’s well-known Content ID system but focuses on identifying a person’s likeness in AI-generated videos. If the system detects a potential match, individuals can review the content and request its removal if it violates YouTube’s privacy policies.

However, detection does not automatically mean the content will be removed. YouTube says it will continue balancing identity protection with free expression. This means that content such as satire, parody, or material serving the public interest may still remain on the platform.

Parent-managed accounts on WhatsApp 

To strengthen the safety of pre-teen users on the platform, WhatsApp is introducing parent-managed accounts. 

This helps families stay connected while also keeping the younger users safe. It allows parents to oversee their child’s digital interactions. Parents can now link a child’s device to their own. By doing this they gain the power to decide who can message the account and which groups the child can join. 

All safety settings and privacy controls are gated by a secure “Parent PIN” on the managed device, ensuring kids can’t bypass these protections.

View on Threads

To Conclude…

And with that whirlwind of updates, we’re wrapping up KIMP’s Picks’ March 13th Edition. Keeping track of the constant stream of updates throughout the week can get overwhelming. But at the same time, you can’t really afford to miss what’s changing either.

That’s exactly why this weekly roundup exists. We sift through the updates, announcements, and releases so you don’t have to spend hours stuck in an endless scroll trying to catch up.

So stay tuned. We’ll be back before you know it with the next round of updates worth knowing. Or to get these updates delivered straight to your inbox, subscribe to The KIMP Scoop newsletter.