Disinformation

What to do when the ‘public good’ of information goes bad

Character.ai is one of the most popular consumer AI apps after ChatGPT, with roughly 20 million monthly active users.

AI’s next smart move might be scrapping the chatbot

Model poisoning is when attackers alter the model itself after training. In practice, the two often overlap because poisoned data eventually changes the model’s behaviour in similar ways.

The rising threat of ‘AI poisoning’

For Singapore as a whole, regional collaboration is essential. Sharing intelligence, detection methods and best practices with Asean neighbours can help us anticipate tactics before they reach our shores.
THE BROAD VIEW

Beyond politics, disinformation and hybrid threats strike at Singapore’s economy

A makeshift memorial for conservative activist Charlie Kirk in Phoenix, Arizona. While partisan disagreement is as old as the republic, today’s polarisation represents a qualitatively different phenomenon.
THINKING ALOUD

The deeper divide: America’s political polarisation crisis

Zuckerberg might talk a good talk about how he is no longer going to be caving in to government demands, but he still is caving in — just to different ones.
THE BOTTOM LINE

The cravenness of Mark Zuckerberg

Under the guise of free speech, Meta CEO Mark Zuckerberg does not have to worry about whether any of its content is factually correct.

Meta’s fact-checking reversal lets Zuckerberg drop the charade

Falsehoods and disinformation can derail companies as online fakery can morph and multiply quicker than ever before.
THE BOTTOM LINE

The disinformation storm is now hitting companies harder