The reliability of news has always been a GP question. What is new is the supply side: generative AI now makes a convincing fake video cost a few dollars and an hour, which moves the issue from 'can we trust the press' to 'can we trust our own eyes'.
A misinformation question rewards a student who can separate two different worries: that people are lied to, which is old, and that the tools for lying have become cheap, fast and near-perfect, which is new. Argue what AI changes and what it does not, and ground it in real 2026 cases, not vague dread.
The reliability of information is a recurring media strand. It surfaced in 2021 as whether the news today is reliable, and the wider cluster on social media's power runs alongside it, from the 2019 question on social media outweighing politicians to the 2024 stem on everyone talking and no one listening. The theme is steady; the technology under it has changed completely.
The current-affairs hook is hard for an examiner to miss. The World Economic Forum named misinformation and disinformation the top short-term global risk in both its 2024 and 2025 reports, and kept it among the leading risks in 2026. When the people who survey global elites for a living rank lies above war and recession, the issue is in the air.
Framed honestly, SEAB sets the paper, and nobody outside it knows the wording. A reliability-of-information stem recurs, and the deepfake era is its freshest and most examinable form. No one can tell you the question. We can tell you the issue is live.
set at A-Level most recent appearance. The reliability-of-news angle last set in 2021; the wider media-power cluster recurs almost every cycle. The AI turn has sharpened it.
These questions test whether you can tell a difference in scale from a difference in kind. Propaganda and forgery are ancient, so a script arguing 'lies have always existed' has said nothing. The arguable claim is whether cheap, convincing, instant AI fakes break something that older lies left intact: the default that a photo or a recording is evidence of what happened.
They also test resistance to your own panic. The strongest answers grant that societies have absorbed every previous information shock, from the printing press to the tabloid to the doctored photo, before pressing on what might be different this time. A doom-only essay and a calm-down-it-is-fine essay both dodge the real argument.
Two distinct, defensible routes through the question. A strong script commits to one and uses the other as the concession it answers, rather than sitting on the fence.
When a believable fake costs almost nothing to make and seconds to spread, the damage is not one false story but the slow loss of any shared baseline of fact, which democracy needs to function.
The panic outpaces the evidence: people are more sceptical than the doom story assumes, verification tools and norms are catching up, and the named deepfakes so far have been caught and corrected rather than decisive.
Singapore's Protection from Online Falsehoods and Manipulation Act, in force since 2019, lets the government order corrections or removal of online falsehoods, and the 2024 elections law adds a specific ban on candidate deepfakes during a campaign.Source: Singapore MDDI, as of 2026-06
The 2024 to 2026 election cycle produced documented political deepfakes beyond Ireland, including a fabricated AI video of a US Senate candidate in the 2026 American midterm campaign, so the threat is recurring rather than a single scare.Source: CNN Politics, as of 2026-06
The sharper danger is less any single fake than the 'liar's dividend': once audiences know that anything can be faked, genuine recordings can be waved away as fakes, which corrodes evidence itself.Source: established disinformation-research framing, as of 2026-06
The old question was whether the news was lying to you. The new one is whether you can still tell the difference, when the fake costs a few dollars and looks exactly real.What AI changes about the question
Book a free trial lesson and see how we teach GP from the exam backwards, building the two-stance habit on issues exactly like this one.