But what if the chatbot has simply inherited somebody else’s version of events?
In January, I wrote about the risk of Wikipedia’s editorial biases finding their way into AI. A recent piece from HonestReporting brings that question closer to home. What happens when a dispute over the wording of a Wikipedia article becomes part of the answer a chatbot gives you today?
HonestReporting points to a Toronto event advertised as a “Free Palestine Wiki Edit-a-Thon.” Organisers offered to train participants to edit Wikipedia and then work together on articles. There is nothing inherently wrong with teaching people to edit an encyclopedia. But when an event is organised around an explicit political cause, it’s fair to ask how the resulting edits will be checked for balance.
Wikipedia has already had to referee disputes over articles on Israel and the Palestinians. In January 2025, its Arbitration Committee barred eight editors from working on the topic. Reporting at the time identified six as pro-Palestinian and two as pro-Israel. The bans show that Wikipedia has a way to deal with disruptive editing. They also show how fiercely people fight over what goes into an encyclopedia many readers take at face value.
There are two ways a source can influence an AI answer. Its text can be part of the material the model learned from during training. Or an AI tool with web access can find the page when you ask your question and use it to compose a response. HonestReporting’s piece highlights that second route. It gives us reason to scrutinise the answers, although tracing any one answer to a particular edit takes more evidence.
An outright falsehood is often the easiest problem to spot. Framing is harder. Which event gets the first paragraph? Whose account is presented as fact and whose is introduced as a claim? What happened the day before, and did it make it into the summary? Every sentence can be defensible on its own while the overall account leaves a misleading impression.
On Wikipedia, a curious reader can look at the footnotes, the editing history and the arguments on the discussion page. A chatbot can compress all of that into three tidy paragraphs. Unless you follow the links, you may never know there was an argument in the first place.
I use AI and will keep using it. It’s a quick way to get oriented on an unfamiliar subject. But on a disputed question, I want to see where the answer came from. If three citations all lead back to the same original claim, I’ve still got one source. And if the chatbot gives me a conclusion I’m already inclined to believe, that’s an especially good time to open the links.
The Israel example should concern us whatever our politics. The same mechanism can carry someone’s preferred account of an election, a public health controversy or a government’s record. The more people rely on a single answer, the more it matters who shaped the sources behind it.
The next time AI gives you a confident answer on a question people are still fighting over, ask it one more question. What evidence would make you change that answer? Then see whether the evidence it gives you is worth trusting.
The American midterm elections have produced a decisive victory for the Democrats. They have taken control of both the House of Representatives and the Senate, seriously restricting Donald Trump's ability to pursue his legislative agenda.
The headlines have been dramatic. A blue wave. A repudiation of Trump. A president whose second term is already in trouble.
There is no disguising the result. Losing both houses of Congress is a substantial political setback.
But before accepting the prevailing interpretation, two questions deserve attention.
How unusual is this result historically? And why does the discussion of Trump's failures so often overshadow any examination of what his administration has actually achieved?
American presidents have a long history of suffering setbacks in midterm elections. The president's party has lost House seats in 18 of the previous 20 midterms.
Consider the record:
The Senate tells a more serious story. Democrats have flipped at least seven Republican-held seats and taken control of the chamber. That is a significant shift, particularly given the states in which they won.
Losing both houses matters. It changes the balance of power in Washington and gives Democrats the ability to block legislation and pursue investigations.
But historical perspective matters too.
A midterm defeat does not automatically mean that a presidency has failed, any more than a midterm victory guarantees future success.
Much of the post-election discussion has focused on Trump's personality, his controversial statements, the Iran war and the cost of living.
Those are legitimate subjects.
But a serious assessment of his administration should also consider its record.
Border security. The reduction in illegal crossings at the southern border has been dramatic. Government figures show a substantial decline in apprehensions. Whatever one's views about immigration enforcement, the reduction is a significant policy outcome.
Crime. FBI figures show that violent crime fell sharply in 2025, including a substantial decline in murder. The downward trend began before Trump's return to office, so he cannot claim all the credit. Nevertheless, the continuing improvement is an important development that deserves attention.
Government efficiency. The administration has challenged federal spending, cancelled contracts and pursued alleged waste and fraud. Its Department of Government Efficiency has claimed substantial savings.
China, Iran and Venezuela. Trump has pursued a confrontational approach toward governments he regards as strategic adversaries. His supporters point to increased pressure and efforts to constrain their influence.
These are not insignificant areas of government policy.
Yet they have often struggled to compete for attention with Trump's latest controversy.
Perhaps the most misunderstood issue in this election has been the cost of living.
Victor Davis Hanson has pointed to a striking comparison. Inflation reached 9.1% during Biden's presidency, compared with 3.7% in September 2026. Gasoline prices were also exceptionally high in 2022.
But there is a distinction that seems to be missing from much of the political discussion.
Reducing inflation does not reduce prices. It merely slows the rate at which they increase.
Suppose groceries cost $100 before the inflationary surge. After a 20% increase, they cost $120. Reduce inflation to 3%, and the following year they cost $123.60.
The inflation rate has fallen dramatically. Yet the family is still paying nearly 24% more than before.
Those accumulated price increases do not disappear when a new president takes office.
And this is where Trump's economic record needs to be considered more carefully.
The high inflation of the Biden years left American households with a substantially higher cost base. Even if inflation subsequently moderates, wages must grow faster than prices over a sustained period before purchasing power fully recovers.
Recent figures show that this recovery remains incomplete. Nominal wages have increased, but inflation has absorbed much of those gains.
Trump also bears responsibility for developments during his own presidency. Tariffs and the Iran war have contributed to renewed price pressures, particularly for fuel.
Nevertheless, blaming today's entire cost-of-living problem on the current administration ignores the cumulative effect of earlier inflation.
Voters remember what things used to cost. Economists measure how quickly prices are changing. They are not asking quite the same question.
The political problem is that voters do not experience inflation as an annual percentage. They experience it at the supermarket checkout and petrol pump.
A president may be able to slow inflation relatively quickly. Restoring the purchasing power lost over several years is a much longer undertaking.
And there is little electoral reward for explaining that distinction to someone who is still struggling to pay the bills.
There is another explanation for why achievements may receive less attention than Trump expects.
Trump has an extraordinary ability to dominate the news. Unfortunately for him, he does not always dominate it in ways that help his own cause.
His confrontational style, personal attacks, extravagant claims and repeated threats generate enormous coverage.
Critics regard this as evidence of poor judgement. Supporters often see it as a willingness to challenge established political conventions.
Either way, it has consequences.
Consider Iran.
The administration's military campaign and diplomatic pressure deserve assessment against their strategic objectives. But Trump's repeated public threats have also created expectations that are difficult to satisfy.
When a president announces dramatic consequences and those consequences do not immediately follow, the discussion shifts from what has been achieved to whether the president has lost credibility.
The announcement becomes the story, rather than the outcome.
Would Trump have been better served by saying less, allowing his administration's actions to speak for themselves and announcing results only when they were achieved?
We cannot know whether that would have changed the election.
But it is reasonable to wonder whether a more disciplined public presentation would have made it easier for voters to judge his record.
Trump has been a polarising figure since entering national politics.
Some of that hostility reflects genuine disagreements over policy, democratic institutions and presidential conduct. Some reflects his own deliberate provocations.
But there is also a broader question about the way political narratives develop.
Once a politician becomes defined primarily by controversy, even significant policy achievements can struggle to receive sustained attention.
Conversely, every mistake becomes further evidence supporting an established interpretation.
Trump's relationship with much of the American media has been openly adversarial for years. The coverage frequently applies standards that are not applied equally to other politicians.
There is also a generational dimension. Younger voters encounter politics through universities, social media and cultural institutions, where attitudes toward Trump can be particularly negative.
It raises the question whether voters are receiving a sufficiently complete account of what governments do, rather than a continuous negative account of what politicians say.
Trump's administration should be judged on its actual results—including failures—not simply on its president's latest remarks.
The Republicans have suffered a real defeat. Losing both houses of Congress is consequential, and the Democrats have every reason to regard the result as a significant electoral success.
The electorate has also sent a message about affordability, economic confidence and the direction of the administration.
But midterm elections are not final judgments on presidencies.
History shows that presidents can suffer substantial congressional losses without those losses defining their entire record.
Trump has delivered measurable changes in some areas, made contested claims in others, and created political difficulties through his own conduct.
Perhaps the most interesting lesson is that a president can dominate the political conversation so completely that his personality overshadows his policies—even when those policies have had measurable success.
That may be Trump's greatest political strength.
And, as these midterms suggest, one of his greatest weaknesses.
When Governments Delay, Victims Pay, And accountability quietly disappears.
When a government makes a serious mistake, who pays?
Usually, the victims.
And when the government takes twenty or thirty years to acknowledge that mistake, the victims pay again.
They pay through years of uncertainty, legal battles, financial hardship and the frustration of being told that their suffering has not yet been officially recognised.
By the time the government finally apologises, many of those responsible have retired, the politicians have moved on, and some of the victims are dead.
Delay doesn't just postpone justice. It can destroy the possibility of accountability.
A recent video by Dr John Campbell examines a British parliamentary debate on 27 October 2026 about the government's response to the UK COVID-19 Inquiry.
During the pandemic, governments exercised extraordinary powers. They confined people to their homes, closed businesses, disrupted education, restricted travel and family contact, and imposed vaccination requirements on sections of the population.
Some emergency measures were justified. Others remain disputed. But the damage caused by restrictions was real, and the decisions deserve rigorous scrutiny.
COVID vaccines reduced the risk of severe illness, particularly among vulnerable people. They also caused rare but sometimes serious adverse reactions. Those recognised injuries deserve proper compensation.
Britain established a major public inquiry in 2022. Its report on vaccines and therapeutics was published in April 2026.
The Inquiry recommended reforming the Vaccine Damage Payment Scheme. The government responded in October, acknowledging the need for reform but without committing to the recommended changes or a timetable.
Meanwhile, injured people continue to wait.
This is the problem with inquiries that identify failures but leave the consequences to another department, another review or another government.
An inquiry is not accountability. Accountability requires action.
Campbell's video reminded me how frequently serious institutional failures have followed the same depressing pattern.
First comes the harm. Then denial, resistance or delay. Eventually an inquiry establishes the facts. Years later, there may be an apology or compensation.
By then, the people who made the decisions are often long gone.
Here are examples from Britain, Australia, Canada and the United States.
These cases are not identical. Some involved government decisions, others failures of regulation, corporate misconduct or official secrecy. The dates also distinguish different milestones—recognition, disclosure, apology and compensation are not the same thing.
But the recurring problem is unmistakable.
The institutions responsible for addressing the harm can take years or decades to deliver justice.
Consider the British infected blood scandal. Thousands were infected through contaminated blood products. In 2024, the British Prime Minister finally delivered a comprehensive apology.
Or Hillsborough, where families fought for 23 years before an independent panel exposed the extent of the failures and misleading official accounts.
Or Australia's Robodebt scandal, where people were pursued for debts calculated using an unlawful method before the programme was dismantled and a Royal Commission exposed what had gone wrong.
The victims were expected to comply with government decisions immediately.
The government, however, could take years to accept responsibility.
We tend to treat delays in government inquiries as unfortunate administrative problems.
But delay has consequences—and those consequences are not shared equally.
For the victim, every additional year may mean another year without compensation, another year of legal expenses, another year of illness or financial distress.
For the institution, time can be an advantage.
Public attention fades. Documents become harder to locate. Witnesses die. Ministers retire. Officials move to other jobs.
Eventually, responsibility becomes so dispersed that nobody is held personally accountable.
The government may apologise. Taxpayers may fund compensation. But the individuals whose decisions caused the damage may suffer no consequences at all.
The taxpayer pays for the original failure, pays for the inquiry, and then pays the compensation.
The people responsible may pay nothing.
That is a perverse incentive.
We cannot punish governments for every decision that turns out badly. Emergencies involve uncertainty, and honest mistakes are inevitable.
But serious failures demand more than an apology delivered decades later.
When credible evidence of major government wrongdoing or negligence emerges, there should be a prompt, independent investigation with the power to obtain documents and identify who made the relevant decisions.
Victims should have access to timely compensation without being forced into years of expensive litigation.
Inquiry recommendations should have deadlines, publicly identified responsible officials and regular progress reports.
And where serious negligence, deliberate concealment or misconduct is established, those responsible should face appropriate professional, civil or criminal consequences.
Responsibility should not evaporate simply because someone acted on behalf of the government.
The lesson from these scandals is not that governments must never make mistakes.
It is that governments must not be allowed to escape responsibility by taking decades to acknowledge them.
The COVID pandemic provides another opportunity to test whether we have learned that lesson.
We need to know which decisions were justified, which caused avoidable harm, who made them and what consequences should follow.
We also need to ensure that people genuinely harmed by government policies are not left fighting for recognition until it is too late.
History shows what happens when governments are allowed to delay.
Victims grow old. Some die. The responsible officials disappear into retirement. Eventually a new government offers an apology and sends the bill to taxpayers.
That is not accountability. It is accountability by attrition.
And unless we change the system, the next government failure will follow exactly the same path.
