In honor of Global Accessibility Awareness Day on May 15, Google is rolling out a fresh wave of features aimed at making its platforms more inclusive. From enhanced screen readers to better captions and broader language support, the updates span both Android and Chrome—and they’re all powered by Google’s Gemini AI.

Smarter Image Descriptions with TalkBack

Android’s TalkBack screen reader just got a major boost. Using Gemini AI, it can now describe images in detail—even when no alt text is provided. This is a big win for users who are blind or have low vision.

The real shift? Interactivity.

Instead of a one-way description, users can now ask follow-up questions about an image. Curious about the brand of a guitar in a photo? Or want to know what else is in the background? You can ask, and the AI responds. This level of detail extends across the entire screen, making it easier to get context on anything from a product listing to a social media post.

Captions That Catch More Than Just Words

Google’s updated Expressive Captions feature now picks up the little things that standard captions miss—like tone, inflection, and background noise. Elongated words like “nooo,” or sounds like throat clearing and whistling, now show up in captions. It might seem small, but for people who rely on captions, these nuances can make a big difference in understanding mood, tone, or sarcasm.

Chrome Gets Friendlier for Visual Impairments

Over on Chrome, accessibility is getting a practical upgrade.

One of the biggest changes: Optical Character Recognition (OCR) is now supported for scanned PDFs. That means screen readers can finally access and read text that was previously locked in image-based files.

Chrome for Android also gets a new Page Zoom feature. It lets users increase text size without breaking the page layout—a long-overdue improvement for those with visual impairments who’ve struggled with clunky, distorted pages.

Breaking Language Barriers with African Speech Recognition

In a push for more global accessibility, Google is investing in speech recognition for African languages. The company is releasing open-source data for 10 languages to help developers build tools for underserved communities.

This move aims to reduce the digital divide by making voice technology accessible in parts of the world often left out of mainstream AI development.

Accessibility at the Core

These updates signal more than just new features—they reflect a shift in how Google designs its products. By integrating AI at the core of accessibility tools, the company is making digital spaces more usable for everyone.

For millions of users with disabilities, this means better tools to navigate, understand, and interact with the world online. And for developers, it offers new building blocks to create more inclusive tech.

Walmart is bracing for a major shift in how people shop—one where artificial intelligence takes the wheel.

The company is updating its digital platforms to accommodate AI-powered shopping agents. These tools, like OpenAI’s Operator, can browse, select, and purchase products based on your preferences. You don’t need to scroll or click. Just tell the AI what you want, and it handles the rest.

To stay ahead, Walmart is developing its own AI-driven features for both its app and website. These tools already assist with simple tasks like reordering groceries, but the goal is much bigger. Soon, you might be able to type or say something like, “plan a unicorn-themed birthday party,” and Walmart’s AI will generate a tailored shopping list, from balloons to cake mix to party favors.

This kind of hands-free, intuitive shopping experience could change how people interact with retailers. Instead of navigating menus and reading reviews, you’ll rely on AI agents to do the legwork. For busy families or anyone who dreads errands, that’s a game-changer.

But this evolution isn’t just about saving time. It’s reshaping how Walmart thinks about digital retail. Traditional marketing techniques—eye-catching product images, catchy slogans, emotional branding—are aimed at human shoppers. AI agents operate differently. They evaluate text-based product information, compare data points, and make choices based on user input and logic. That forces Walmart to rethink how it writes product descriptions, sets prices, and structures its promotions.

These changes also affect how Walmart interacts with third-party AI agents. The company knows consumers won’t always use Walmart’s own tools—they may rely on assistants developed by tech giants or independent platforms. To prepare, Walmart is building systems that allow outside AI agents to communicate directly with its product databases. This includes sharing user preferences, receiving tailored product recommendations, and completing transactions smoothly, no matter which AI is doing the shopping.

Behind the scenes, this shift demands a complete overhaul of how retail platforms work. It means building trust between retailers and AI agents, ensuring product information is consistent and clear, and creating standards that allow machines to interpret and act on user intent.

Even though the majority of shopping still happens in physical stores, the signs of change are clear. AI shopping agents are gaining traction, especially for routine purchases and online orders. As they become more capable, their role in everyday shopping will grow.

Walmart’s early moves show it’s paying close attention to that future. By investing in AI now—before these agents become mainstream—it’s positioning itself to serve not just the customers of today, but the algorithms that might represent them tomorrow.

In recognition of Global Accessibility Awareness Day, Meta has announced a series of initiatives aimed at enhancing accessibility across its range of products and platforms. These developments focus on providing more inclusive experiences for users with disabilities, leveraging advanced technologies to break down barriers.

Enhanced AI Capabilities on Ray-Ban Meta Glasses

Meta’s Ray-Ban Meta glasses, known for their hands-free functionality, are receiving an update that allows users to customize Meta AI for more detailed responses. This feature enables the AI to provide descriptive information about the user’s surroundings, which can be particularly beneficial for individuals who are blind or have low vision. The update is set to roll out in the U.S. and Canada, with plans for broader availability in the future.

Additionally, Meta is expanding its “Call a Volunteer” feature, developed in partnership with Be My Eyes. This service connects users with sighted volunteers in real-time to assist with everyday tasks. The feature will soon be available in all 18 countries where Meta AI is supported.

Advancements in Human-Computer Interaction

Meta is exploring the use of surface electromyography (sEMG) wristbands to facilitate human-computer interaction, particularly for individuals with physical disabilities. These wristbands detect muscle signals at the wrist, allowing users to control devices even if they have limited mobility due to conditions like spinal cord injuries or tremors. Recent research collaborations, including one with Carnegie Mellon University, have demonstrated the potential of sEMG technology to enable users with hand paralysis to interact with computing systems effectively.

Improving Communication in the Metaverse

To make virtual experiences more accessible, Meta is introducing live captions and live speech features in its extended reality products. Live captions convert spoken words into text in real-time, while live speech transforms text into synthetic audio. These features aim to assist users who have hearing impairments or prefer alternative communication methods. Enhancements include the ability to personalize and save frequently used messages.

Furthermore, developers at Sign-Speak have utilized Meta’s open-source AI models to create a WhatsApp chatbot that translates American Sign Language (ASL) into English text and vice versa. This innovation facilitates communication between Deaf individuals and those who do not understand ASL, using avatars to convey messages in sign language.

Wrap Up

Meta’s ongoing commitment to accessibility reflects its dedication to creating inclusive technologies that cater to the diverse needs of its global user base. By integrating advanced AI and human-computer interaction technologies, Meta aims to empower individuals with disabilities to engage more fully with digital experiences.

Spotify has enhanced its AI DJ feature, allowing Premium subscribers to personalize their listening experience using English voice commands. Previously, the AI DJ generated playlists based solely on users’ listening habits, with limited control over what played next. Now, users can press and hold the ‘DJ’ button in the app, wait for a beep, and speak commands such as requesting specific genres, artists, moods, or even imaginative prompts like “play me some music to soundtrack my life as a movie.”

This update brings more direct control and integrates quirky, vibe-based recommendations similar to those found in Spotify’s AI Playlist beta feature. However, voice commands remain the only way to guide the AI DJ’s selections, which may be inconvenient in quieter or public settings.

The AI DJ feature, introduced in February 2023, uses a synthesized voice to provide a personalized radio listening experience. The voice is modeled after Spotify’s Head of Cultural Partnerships, Xavier “X” Jernigan. Spotify has also launched a Spanish-language version of the AI DJ, expanding its accessibility to a broader audience.

To use the new voice command feature, users can:

• Open the Spotify app and search for “DJ.”
  
• Tap play to launch the AI DJ.
  
• Press and hold the DJ button (bottom right) until a beep.
  
• Speak your request, like: [“Play me chill tracks for a rainy afternoon.”], [“Give me some K-pop with choreography vibes.”], [“Surprise me with indie songs I’ve never heard.”].

If you’re not sure what you want, just tap the DJ button to skip to the next vibe.

This feature is now available in over 60 markets worldwide, offering users a more interactive and personalized music experience.

Imagine sending a message that doesn’t travel through wires or bounce off satellites but instead appears instantly at its destination—no physical journey, just a seamless transfer of information. This isn’t science fiction; it’s the essence of quantum teleportation. And scientists at Northwestern University achieved a significant milestone by demonstrating quantum teleportation over existing fiber optic cables already carrying internet traffic.

Quantum teleportation doesn’t involve moving objects instantaneously from one place to another. Instead, it refers to the transfer of quantum information—the state of a quantum particle—from one location to another without traversing the space in between. This process relies on a phenomenon called quantum entanglement, where two particles become linked, and the state of one instantly influences the state of the other, regardless of the distance separating them.

In the recent experiment, researchers successfully teleported quantum information over 30 kilometers (approximately 18.6 miles) of standard fiber optic cable that was simultaneously transmitting conventional internet data at 400 gigabits per second . This achievement marks the first time quantum teleportation has been demonstrated over existing internet infrastructure without the need for specialized, dedicated channels.

Transmitting quantum information over busy internet cables presents significant challenges. Quantum states are incredibly delicate and can easily be disrupted by noise from other data transmissions. To overcome this, the Northwestern team utilized a less congested wavelength of light, known as the O-band, for the quantum signal. They also implemented narrow spectro-temporal filtering and multi-photon coincidence detection to protect the quantum fidelity from noise, ensuring the quantum information remained intact during transmission.

This approach allowed quantum and classical communications to coexist on the same fiber optic cable, demonstrating the feasibility of integrating quantum communication into our current internet infrastructure.

The successful demonstration of quantum teleportation over existing internet cables is a significant step toward the development of a quantum internet. Such a network would enable ultra-secure communication, as any attempt to intercept quantum data would immediately alter its state, revealing the presence of an eavesdropper. This has profound implications for fields requiring high levels of security, such as banking, healthcare, and national defense.

Moreover, integrating quantum communication with existing infrastructure could accelerate the deployment of quantum networks, making them more accessible and cost-effective. As quantum computers become more prevalent, the ability to transmit quantum information reliably and securely will be crucial for harnessing their full potential.

While the concept of teleportation often conjures images of science fiction, the reality of quantum teleportation is firmly rooted in scientific research and experimentation. The recent breakthrough by Northwestern University researchers demonstrates that quantum communication can be achieved using the same infrastructure that supports our current internet, bringing us closer to a future where quantum networks are an integral part of our digital landscape.

As we continue to explore the possibilities of quantum technology, each advancement brings us closer to a new era of communication—one where information transfer is instantaneous, secure, and seamlessly integrated into our everyday lives.

Microsoft has announced a new security feature for its Teams platform aimed at preventing unauthorized screen captures during meetings. Set to roll out globally in July 2025, this “Prevent Screen Capture” feature will be available on Teams desktop applications for both Windows and Mac, as well as on mobile applications for iOS and Android.

According to Microsoft’s Microsoft 365 roadmap, the feature is designed to address concerns over unauthorized screen captures during meetings. If a user attempts to take a screenshot, the meeting window will turn black, thereby protecting sensitive information shared during the session.

To further safeguard content, users joining meetings from unsupported platforms will be automatically placed in audio-only mode, ensuring that sensitive visuals are not exposed.

While this feature enhances security, Microsoft acknowledges that it cannot prevent all forms of content capture.

For instance, individuals could still use external devices, like cameras, to photograph the screen. Details regarding whether this feature will be enabled by default or require activation by meeting organizers or administrators have not been disclosed.

This development aligns with broader industry trends focusing on privacy and data protection. For example, Meta recently introduced an “Advanced Chat Privacy” feature in WhatsApp, which blocks attempts to save shared media and export chat content in private and group conversations.

In addition to the screen capture prevention feature, Microsoft plans to roll out other updates in June 2025, including town hall screen privilege management for Teams Rooms on Windows, interactive BizChat/Copilot Studio agents in meetings and one-on-one calls, and a Copilot feature to generate audio overviews of transcribed meetings.

These enhancements reflect Microsoft’s ongoing commitment to improving security and user experience within its collaboration tools.