Know better. Do better.

When Donald Trump gets back into the White House, he's expected to prioritise the deportation of undocumented immigrants- one of his major campaign promises.

As we've reported at Context, a deportation regime will rely on all sorts of tech that's already in place: from risk algorithms that decide who is forced to wear an ankle monitor, to data-brokers that can help law enforcement pinpoint the location of an undocumented migrant.

More recently, the Department of Homeland Security (DHS) which houses immigration enforcement authorities, has been experimenting with AI tools for a range of tasks, including a system that could automatically flag asylum applications as fraudulent.

Groups that advocate for migrant rights have been raising the alarm about this technological infrastructure for years, saying these big-data systems give authorities too much power and could lead to abuse.

The Trump campaign said it would look to deport around a million people a year, but experts we spoke to were divided about how realistic a mass deportation regime would be, given the legal and bureaucratic constraints. 

We spoke to Julie Mao, a lawyer with the immigration rights organisation Just Futures Law. She predicted a Trump administration would be able to rely on tech tools to supercharge the deportation process.

"With their surveillance tools and the network of data brokers, they could easily populate long lists of current locations of immigrants they want to deport," she said. "They are ready to go."

Africa

Kim Harrisberg, South Africa Correspondent

While Elon Musk's Starlink continues to spread across Africa, the international telecommunications provider paused subscriptions in five Kenyan regions, including the capital Nairobi, due to high demand and capacity issues, news site Tech Cabal reported.

This pause has raised questions around Starlink's capacity to cope with internet demands in densely populated urban areas.

The number of Starlink users in Kenya has grown tenfold since its launch in July 2023 as the cost of internet access kits and monthly plans were reduced to around $15 a month, compared to the purchase cost of $350.

Latin America

Diana Baptista, Data Journalist

At least 75.4 million people – 34% of the population – are under surveillance by opaque facial recognition systems used for public security in Brazil, according to a recent report by Panóptico, a project that monitors facial recognition in the country, and the Laboratory of Public Policy and Internet (LAPIN).

Of the 264 systems operating in the country, most lack proper regulation and transparency mechanisms. Municipalities and companies fail to disclose basic information, such as where the facial recognition cameras are located, who manages the data they generate and where the data is stored.

The research found that 55% of the systems operate without information on the company behind them; 70% of the projects have no information on how the systems were acquired, and 80% do not have impact evaluations.

Human rights groups have criticised facial recognition systems in Brazil for biases that have boosted racial profiling and false positives in multiple Brazilian municipalities.

Asia

Annie Banerji, South Asia Correspondent

Australia's government plans to ban social media for children under 16, with legislation expected by late 2025.

Prime Minister Anthony Albanese said social media harmed children's physical and mental health. 

The country is testing age-verification methods, including biometrics, to enforce the ban. Australia's policy is one of the toughest globally, with no exemptions for parental consent or pre-existing accounts, unlike France, which banned under-15s to use apps generally, but had a carve-out for young users with parental consent.

Europe

Adam Smith, UK Tech Correspondent

Esther Ghey, mother of murdered teenager Brianna Ghey, criticised the chatbot platform Character.AI for allowing users to create digital avatars of her late daughter and other individuals, including suicide victim Molly Russell. 

The app failed to prevent the creation of chatbots mimicking Brianna and others, including using personal photos and details. 

A spokesman for Chartacter.AI told The Telegraph newspaper that it moderates proactively and in response to user reports.

The company was previously embroiled in a scandal after an American teenager took his own life after talking to a Daenerys Targaryen chatbot on the site.

A spokesperson for the website told The New York Times that it would be implementing safety features for younger users, including a time limit feature, a warning message that users should "treat everything (the bot) says as fiction".