原田英語の時事英語ジュニア|英語ニュース ワークシート(高校生レベル)
Class ____ No. ____ Name ____________________
偽の記憶とは?AIで加工した写真から作った動画で、まちがった記憶が約2倍に
Photos Edited by AI Can Plant False Memories, Study Finds
1Seeing photos that have been edited by AI can make people remember details that were never there, researchers at the MIT Media Lab in the United States have found. The study was first shared online in 2024, and U.S. news outlets reported on it this month. One of the authors is Elizabeth Loftus, a psychologist known for her work on false memories.
2In the experiment, 200 adults in the United States, aged 20 to 73, looked at 24 photos for two minutes. The photos included scenes from daily life, the news, and history. After playing Pac-Man for two minutes, the participants were split into four groups of 50. Each group was shown a different version of the images, which were described as enhanced.
3One group saw the original photos again. The others saw photos edited with AI, short AI videos made from the original photos, or AI videos made from the edited photos. The edits included changing a person's expression, changing the time of day, and adding a stop sign.
4The participants then answered 24 questions about the original photos. The group that watched AI videos made from edited photos had the most false memories, about 2.05 times as many as the group that saw the originals again. These participants were also about 1.19 times as confident in their false memories.
5The researchers said the results raise ethical, legal, and political concerns. They also said the same effect might one day be used in a helpful way, such as helping people see painful memories in a new light.
Words 単語
| editedB2 | 動 | 加工された、手を加えられた(edit の過去分詞)changed to make something different or better |
| detailsA2 | 名 | 細かいところ(detail の複数形)small facts or parts of something |
| outletsB2 | 名 | (報道の)媒体、メディアcompanies that share news, such as papers or TV |
| psychologistA2 | 名 | 心理学者a scientist who studies the mind and actions |
| participantsB1 | 名 | 参加者people who take part in an activity |
| versionB2 | 名 | 版、形one form of something that is a bit different |
| expressionA2 | 名 | 表情the look on a person's face |
| confidentA2 | 形 | 自信のあるfeeling sure that you are right |
| ethical外 | 形 | 倫理の、道徳的なabout what is right and wrong |
| painfulB1 | 形 | つらい、痛いmaking you feel hurt or very sad |
単語のレベル:CEFR-J(『CEFR-J Wordlist Version 1.6』 東京外国語大学投野由紀夫研究室. (URL: https://www.cefr-j.org/download.html より 2026年8月ダウンロード))
1. Reading Quiz 本文の内容に合うものを選ぼう
Who is Elizabeth Loftus?
A. A psychologist known for her work on false memoriesB. A photographer at a newspaperC. A video game designerD. A student who took part in the study
How many people were in each group?
A. 24B. 200C. 50D. 73
Which of these was one of the edits made by AI?
A. Making the photos black and whiteB. Removing all the peopleC. Adding musicD. Changing the time of day
Which group had the most false memories?
A. The group that saw the original photos againB. The group that watched AI videos made from edited photosC. The group that saw AI videos made from the original photosD. The group that played Pac-Man the longest
What helpful use did the researchers mention?
A. Making better selfiesB. Finding lost phonesC. Helping people see painful memories in a new lightD. Teaching AI to draw
2. True or False 正しければ T、違っていれば F
( ) The study was first shared online this month.
( ) The participants were told the second set of images had been enhanced.
( ) The participants answered 24 questions about the original photos.
( ) People who watched AI videos of edited photos were less confident in their false memories than the group that saw the originals.
3. Fill in the Blanks 下の語から選んで空らんをうめよう(形を変えずに)
versionexpressionpsychologistparticipantseditedoutlets
The study was first shared online in 2024, and U.S. news _____ reported on it this month.
One of the authors is Elizabeth Loftus, a _____ known for her work on false memories.
The edits included changing a person's _____, changing the time of day, and adding a stop sign.
Each group was shown a different _____ of the images, which were described as enhanced.
After playing Pac-Man for two minutes, the _____ were split into four groups of 50.
Seeing photos that have been _____ by AI can make people remember details that were never there, researchers at the MIT Media Lab in the United States have found.
4. Word Order 語を並べかえて正しい文にしよう
( history / photos / scenes / The / and / life, / included / news, / the / from / daily ) .
( One / again / original / group / photos / saw / the ) .
( about / questions / the / 24 / photos / original / answered / then / participants / The ) .
5. Think and Write 英語で答えてみよう
Should social media apps be required to label photos and videos edited by AI? Why or why not?(SNS のアプリは、AI で加工した写真や動画に「AI 加工」と表示することを義務にすべきだと思いますか?理由も考えよう。)
If AI can change our memories, how should we use photos when we study history?(AI が記憶を変えられるなら、歴史を学ぶとき写真をどう使えばよいでしょうか?)

Answers 解答
1. (1) A (2) C (3) D (4) B (5) C
2. (1) F (2) T (3) T (4) F
3. (1) outlets (2) psychologist (3) expression (4) version (5) participants (6) edited
4.
- The photos included scenes from daily life, the news, and history.
- One group saw the original photos again.
- The participants then answered 24 questions about the original photos.
全文訳
1AI で加工された写真を見ると、人は実際には無かった細かいところまで覚えていると思いこむことがあると、アメリカの MIT メディアラボの研究者たちが明らかにしました。この研究は2024年に初めてネットで公開され、今月アメリカの報道機関が取り上げました。著者の1人は、偽の記憶の研究で知られる心理学者のエリザベス・ロフタスさんです。
2実験では、アメリカに住む20〜73歳の大人200人が、24枚の写真を2分間見ました。写真には、ふだんの生活やニュース、歴史の場面がうつっていました。2分間パックマンで遊んだあと、参加者は50人ずつ4つのグループに分けられました。グループごとにちがう形の画像が見せられ、それらは「きれいにした画像」と説明されていました。
31つのグループは、もとの写真をもう一度見ました。ほかのグループは、AI で加工した写真、もとの写真から AI が作った短い動画、加工した写真から AI が作った動画のどれかを見ました。加工には、人の表情を変える、時間帯を変える、止まれの標識を足す、といったものがありました。
4そのあと参加者は、もとの写真について24の質問に答えました。加工した写真から作った AI 動画を見たグループは、まちがった記憶がいちばん多く、もとの写真をもう一度見たグループの約2.05倍でした。このグループの人たちは、まちがった記憶への自信も約1.19倍でした。
5研究者たちは、この結果は倫理・法律・政治の面で心配な点を生むと述べました。一方で、同じ効果がいつか、つらい記憶を新しい見方でとらえ直す手助けのように、役に立つ形で使えるかもしれないとも述べました。
文法ポイント:倍数表現(~ times as many as ...)
「…の~倍の数」を表す。本文の about 2.05 times as many as the group that saw the originals again(もとの写真を見たグループの約2.05倍)がその例。数えられない量なら as much as、性質なら as confident のように形容詞を入れる。
出典:KPRC Click2Houston、arXiv(MIT Media Lab の研究者ほか)、MIT Media Lab ほか(記事は原田英語の時事英語ジュニアのオリジナル)|https://www.haradaeigo.com/jijieigo-junior/news/2026-10-26-ai-edited-photos-false-memories/