Skip to main content
lately

Welcome back to Lately, The Globe’s weekly tech newsletter. I hope everyone had a restful holiday. If you have feedback or just want to say hello to a real-life human, send me an e-mail.

In this week’s issue:

⛔ X limits access to AI image generation after backlash over sexualized deepfakes

✨ The genius and pitfalls of vibe-coding

👩🏻‍💻 Ottawa halts visa program for startup entrepreneurs

🤖 The best of CES


DEEPFAKES

X limits access to Grok image generation after barrage of sexualized deepfakes

Open this photo in gallery:

Women speak out against nonconsensual deepfakes on X.Tita Barros/Reuters

Over the past two weeks, a flood of explicit, AI-generated images of real people have proliferated on the social media platform X, triggering condemnation and investigations from governments around the world into the Elon Musk-owned company.

The images were generated with Grok, the platform’s built-in AI chatbot that allows users to reply to posts on X with questions or requests. Since late December, Grok has created tens of thousands of explicit images upon user request, altering real photos of people without their consent to depict them in sexual poses or minimal clothing. Earlier today, X began limiting image generation to users with paid subscriptions.

“Nudify” apps and websites, which produce sexualized deepfake images of real people using generative AI, are not a new phenomenon. But Grok – which has looser restrictions than other chatbots and is seamlessly integrated into X – has pushed deepfakes into the mainstream. For online safety advocates, the issue represents a perfect storm: As technology advances, governments are failing to keep up with regulating the platforms where this technology is prevalent.


IMMIGRATION

Immigration Department shelves visa program for foreign entrepreneurs

In late December, the Immigration Department halted a settlement program for foreign entrepreneurs that has faced accusations of misuse and has processing times of more than 10 years. The government said it’s looking at establishing an alternative with stricter rules for who can participate. However, immigration lawyers warn that closing the Start-up Visa Program, which was created in 2013 to boost jobs, could actually create more uncertainty and lead to foreign investors choosing Europe or the U.S over Canada.


ARTIFICIAL INTELLIGENCE

Vibe-coding brings computer programming to everyone

With the rampant advancement of AI tools, a new style of coding has become all the rage, where instead of toiling over lines of code, you describe what you want in plain language and let the AI do all the actual code-writing. The approach is known as vibe-coding – a term coined by Andrej Karpathy, the former director of AI at Tesla. (It was also one of The Globe’s 2025 tech words of the year.)

Among some programmers and Redditors, vibe-coding is derided as a realm for charlatans who go on to produce buggy software full of security risks. But large AI companies, such as OpenAI and Anthropic, have long focused on building tools to speed up software development, and code-writing startups such as Cursor and Lovable are now worth billions of dollars.

As tech reporter Joe Castaldo writes, vibe-coding platforms aren’t all that different from any other generative AI app: They can be impressive and useless, accurate and error-prone, solve your exact problem and lead you astray, all while carrying the potential to completely change how software is made. Read the full story here.

What else we’re reading this week:

A global crackdown on online gambling (Rest of World)

Meet the high school kids cutting $25,000 venture checks in Silicon Valley (The Wall Street Journal)

How OpenAI used aggressive discounts to dominate AI in universities (Los Angeles Times)

Adult Money

GADGETS

The best of CES

Open this photo in gallery:

The LG CLOiD robot, one of many humanoid robots launched at CES.Steve Marcus/Reuters

This week, tech companies rolled out their newest gadgets and gizmos at CES, the annual trade show in Las Vegas organized by the Consumer Technology Association. Sure, Nvidia and Intel had important chip announcements, but the most exciting innovations were the laundry-washing humanoid robots, gaming laptops with rollable screens that can expand two inches, vacuum bots with legs to tackle stairs, and a “longevity station” – aka a very fancy scale – that can measure over 60 biomarkers.

But the product that attracted the most eyeballs was Lego’s smart bricks, which contain sensors that enable them to produce lights and sounds when used in unison. The strangest product? An AI desk companion, which is essentially a hologram of an avatar that can sit on your desk.

Culture radar

MUSIC

The music industry is going all-in on licensed AI – but musicians are less excited

Open this photo in gallery:

Music execs love AI. Musicians don'tIllustration by Kyle Smart

The rise of generative AI spurred, at first, near-universal condemnation from creative industries. Actors were worried their likeness would be used in movies and TVs. Writers were furious that AI models have scraped and ingested their copyrighted works. Music companies Warner and Universal began to take AI music startups to court.

But as arts reporter Josh O’Kane writes, in the past few months, the dynamic has shifted in the music industry. Both Warner and Universal have settled their legal battles, and have even begun licensing their recording and publishing catalogues to those same AI music generators. Canada’s songwriting royalty-collection organization SOCAN will start accepting compositions that have been partly generated with AI.

Yet, many working musicians continue to see AI as an existential threat, and there is a great deal of skepticism over how much money will actually be returned to artists. There has always been a chasm between musicians and the music industry, but now it’s wide open. Read the full story here.

Follow related authors and topics

Authors and topics you follow will be added to your personal news feed in Following.

Interact with The Globe