Skip to content

WTC Wrap: 16 May 2026

This week: Ten executions in under five months, Singapore's Universal Periodic Review, thoughts on DRC and drug supervision, and musings on the embrace of AI.

Hello from KL! This morning I'm facilitating a workshop on the media, ethics, and the death penalty, organised by the Anti-Death Penalty Asia Network (ADPAN). Apart from that, I'm really looking forward to burrowing into the hotel bed to read and sleep. Especially sleep. But maybe also some episodes of Ted Lasso.


(1)

Singapore carried out its 10th execution of 2026 yesterday morning. It's a staggering pace of killing.

On Tuesday, during Singapore's Universal Periodic Review at the UN Human Rights Council, a total of 41 countries made recommendations related to the death penalty, such as urging for a moratorium on executions and respecting the right to life. In its press statement on the session, the Ministry of Foreign Affairs reduced this high number of recommendations (even more than the death penalty recommendations made during the last cycle) to a quick and vague reference to "recommendations by several States on the abolition of capital punishment".

Singapore's delegation, led by Minister of State for Health and Digital Development and Information Rahayu Mahzam, insisted that our criminal punishment policies are evidence-based and in line with international law—even though international standards make it clear that "[in] countries which have not abolished the death penalty, capital punishment may be imposed only for the most serious crimes, it being understood that their scope should not go beyond intentional crimes with lethal or other extremely grave consequences". If you're interested, I found a 2013 paper that begins with discussion of this "most serious crimes" limitation and makes clear that the mandatory death penalty for drug offences—which is what Singapore has—does not meet this standard.

"The use of capital punishment in our criminal justice system is not a decision we have taken lightly, and we do so with a heavy heart," Rahayu told the Human Rights Council. I'm glad I'd already gone to bed by this point and didn't see her say this live, or I might still be crawling on my hands and knees, swearing, trying to locate the eyeballs that rolled out of my head. Hanging 10 people in under five months, taking the position that people can be executed despite being party to ongoing legal proceedings or formal complaints against their previous lawyers, and reviewing policy to reduce notice periods for some prisoners is not "heavy heart" behaviour.


(2)

People who use drugs who surrender themselves to the authorities the first time will no longer be sent to Drug Rehabilitation Centres. Instead, they'll be put on drug supervision with compulsory case management. This won't apply to people for whom it's not their first time, or who are already being investigated by the police.

Given that DRC is essentially prison, I'm glad if fewer people have to experience it. But if the drug supervision that Edwin Tong, the minister for law and second minister for home affairs, is referring to is the current drug supervision regime that we have, it's worth noting that it's punitive and dehumanising, too. Research on drug supervision and desistance in Singapore has also pointed to how such surveillance doesn't help people turn things around if they've been using drugs but want to stop. Here's an excerpt from the brief summary published on NUS's website about the paper I linked above:

Through qualitative research involving interviews with 16 current and former drug supervisees and desistors, the article reveals how surveillance-oriented supervision erodes personal capital, undermines motivation, and leads to non-compliance among supervisees. Specifically, the authors highlight the contradictions of the erratic urine testing regime post-release and its counter effects of various dimensions of desistance, namely act, identity, and relational desistance, among recovering drug abusers.  

While the main objective of urine testing is to strengthen deterrence against relapse and to reduce recidivism, consistent testing has led to inadvertent outcomes. Supervisees often resort to methods to defraud the tests, such as flushing their systems or tampering with their urine specimens. Routine testing also disrupted offenders’ abilities to establish themselves in new employment roles and perform familial roles effectively, hence reducing identity desistance. Additionally, the demands of urine supervision strain relationships, limiting relational desistance. Consequently, prolonged supervision hinders desistance efforts, perpetuates a sense of limbo between being an “offender” and “free” individual, and increases recidivism rates. Statistics indicate a 24.3% recidivism rate for drug offenders at the 2-year mark and a worrying 40% at the 5-year mark. For long-term drug abusers, the recidivism rate is almost 60%. These findings suggest re-evaluating models of intervention to ensure that they genuinely support desistance among released drug abusers.

Ultimately, I'm not sure how many people would surrender themselves to the authorities for drug use. It's still a very punitive environment with serious repercussions like social stigma and threats to livelihood. The orientation towards criminalisation and punishment just constantly undercuts any attempt at providing care.



(3)

I've been reading Karen Hao's Empire of AI recently and finding it clarifying and thought-provoking. It's helped me wrap my head around what deep learning and neural networks and Large Language Models mean, and to see more clearly the gulf between what this technology actually is and the lofty promises of "intelligence" and solutions for all sorts of things in our lives and the world.

It's also making me think more about Singapore's embrace of AI. An Economic Strategy Review committee recently made recommendations for Singapore's development into an AI-empowered economy, saying that Singapore should position itself as "a trusted hub where AI solutions are developed, tested and deployed to tackle real-world problems at scale". AI tools are also being incorporated into schools; Minister of Education Desmond Koh has given reassurances that they're going to be age appropriate and will prep kiddos "to harness AI to benefit their learning, critically evaluate AI output and guard against risks such as cognitive offloading". That said, I can't help but wonder: how much training and education and understanding do the teachers have about AI, its ethical pitfalls, the power structures it perpetuates, and its limitations?

Meanwhile, the government has also been trying to reassure Singaporeans that we won't all be left in the dust by the rush to AI. Parliament has adopted a "no jobless growth" motion, with Minister for Manpower Tan See Leng saying: "The Government cannot protect every job, but will certainly do our best to support and protect every worker."


Thank you for reading! As always, feel free to forward this weekly wrap to anyone you think might be interested, share the web link on social media, or spread the word about this newsletter!