AI

HitPaw has rolled out major updates to FotorPea and VikPea. The updated photo editing tool, FotorPea 5.0.0, introduces a new feature called AI Replace, which lets users modify images with plain text instructions. This can be used to switch backgrounds, adjust hairstyles or clothing, make corrections directly within selected areas, and so on.

The face restoration system has been tweaked in this new release, with the CodeFormer model producing details closer to natural facial features.

Global influencer marketing spend is forecast to reach $32.5 billion in 2025, with much of the new growth coming from AI influencers, according to a new case study by UNmiss.

From polished AI Instagram avatars to TikTok personalities who never sleep, brands are increasingly testing whether audiences care if creators are real or virtual.

A new survey from ABBYY highlights the hurdles businesses face when adopting generative AI, revealing that many companies need additional AI tools to get the results they want. The report, State of Intelligent Automation: GenAI Disillusionment and AI Wishlist, points to training, integration and governance as major sticking points.

The research, carried out by Opinium, shows that 31 percent of business leaders find training models more difficult than expected. Another 28 percent say GenAI tools are hard to integrate, while 26 percent cite governance problems. Misuse by staff was flagged by 21 percent of respondents, suggesting that some deployments lack proper oversight.

Microsoft has released a new Canary build of Windows 11, and among the additions is the arrival of AI actions in File Explorer. The company implores users to “leverage[e] the power of AI” to summarize documents and edit images directly from File Explorer new AI Actions context menu entry.

For Microsoft, this is the highlight of Windows 11 Insider Preview Build 27938, but there is much more to this release. There are, of course, lots of bug fixes but there are also new clock options including the ability to show seconds.

Rabbit has announced rabbitOS 2, a brand new operating system update for its r1 AI companion device, redesigning the interface and expanding its AI features.

The update, released today, brings a colorful card-based layout and a new system called creations that lets users build tools, games, and experiences directly on the device by talking to it.

There is much to get frustrated about in relation to AI, not least of which is that there are so many AI tools to choose from. But there is also the fact that each AI app and service has its own limitations, and it can be difficult – or impossible – to know what these limits are until you hit them.

It has been this way with Google Gemini. Google has, until now, been a little fuzzy when it comes to telling users about limits. Being told that the company “may have to cap how much you can use some features” is not very helpful, certainly not very specific. But now Google has published much more detailed information to help users know where they stand.

Timekettle has launched the W4 AI Interpreter Earbuds which combine bone-conduction voice sensors with AI translation models to provide near real-time multilingual communication.

The W4’s bone-voiceprint sensor filters out background sounds, improving clarity in crowded or noisy settings and cuts audio leakage, offering more privacy.

Truecaller has announced an AI-powered expansion of its caller ID service, aiming to reveal not only who is on the other end of the line, but also why they are phoning you.

The company says the new feature brings real-time context to incoming calls, with the goal of reducing uncertainty for its 450 million global users. The update comes as scam calls and fraud attempts continue to skyrocket worldwide.

The inescapability of AI continues, and Meta is at the forefront. WhatsApp has already liberally coated itself in features driven by artificial intelligence, and now there are more.

This time around, things are fairly innocuous. The chat platform has now added the option of using AI backgrounds within video chats.

Alibaba has released a new open-source speech-to-video model capable of generating animated digital humans from a single portrait and an audio clip. The tool is aimed at content creators and researchers who are looking for a way to produce lifelike avatars capable of speaking, singing, or performing.

The Wan2.2-S2V release builds on Alibaba’s Wan2.2 video generation series. By becoming open-source, the company is offering developers a system that can animate portraits across different perspectives including close-up, bust, and full-body shots.

Nearly 40 percent of organizations admit they are underprepared for AI-driven threats such as automated attacks, deepfake-based videos, and voice scams, according to new research from LevelBlue.

The new findings show that while awareness of these dangers is growing, many companies remain vulnerable and lack confidence in their ability to defend against them.

LinkedIn says professionals are turning to their colleagues and personal contacts for advice at work, not search engines or AI tools.

New global research released by the company suggests 43 percent of workers rank personal networks above other options when they need clarity, with two-thirds saying that peers help them make decisions more confidently.

Elon Musk ‘s xAI startup is suing both OpenAI and Apple for alleged anticompetitive behavior relating to artificial intelligence.

Musk is focusing on these two firms because of the partnership and collaborative arrangements that exist between them. OpenAI has an agreement with Apple to integrate its ChatGPT tool into iOS and macOS. This is something that troubles xAI.

Many people turn to AI to improve their writing, their photos, or their videos. YouTube has been doing the same, quietly enhancing the look of users’ uploaded videos through the use of artificial intelligence.

So, what is the problem? Many creators are upset not so much that their videos have been “enhanced” (although there is debate over whether what YouTube has done can actually be seen as an improvement) but that it was done without consent.

Cloudflare has launched new Zero Trust tools in its Cloudflare One platform to help businesses adopt artificial intelligence securely at scale. The features aim to give organizations better visibility into how AI is used, as well as protection from unapproved applications, and controls to safeguard sensitive data, all within a centralized system that can support distributed teams.

Generative AI is now being used by employees across departments, from finance to design, to speed up tasks and create new applications. While this can, and does, help improve efficiency, adoption often happens without security oversight.