2026年11月9日（月）マイページ
← 記事にもどる 中学生版高校生版 Word

「印刷する」→ 送信先で「PDFに保存」を選ぶと PDF になります。解答は最後のページです。

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 ____________________

生成AIで力はつく？弁護士の3か月の実験、AIなしでも伸びていたのはベテランだけ

Study Finds AI Helped Senior Patent Lawyers Build Skills, but Not Beginners

音声・ゲームはこちらから

1Experienced professionals who worked with AI for three months did better on a later task without AI, but beginners did not gain on average, according to a new study. The study was posted as a working paper by the National Bureau of Economic Research and has not yet been peer-reviewed.

2The researchers, including David Autor of the Massachusetts Institute of Technology (MIT) and a team from Google, followed 133 patent lawyers at 11 U.S. law firms. Patent lawyers write the documents that protect new inventions. About two-thirds of the lawyers were randomly given an AI writing assistant built by Google, while the rest worked without one. Expert patent lawyers who did not know which group each person was in scored their work.

3While they had the tool, the AI users wrote better patent drafts than the other group. After three months, however, all of the lawyers were asked to review and mark up a patent draft without AI. Lawyers who had used the tool did better overall, but the entire advantage came from senior lawyers. Junior lawyers, who had been working for less than seven years, showed no gain on average.

4The junior lawyers' results also became more spread out. Compared with juniors who had not used AI, fewer of them got middle scores, while more got poor scores and more got good scores. The authors concluded that basic expertise may be needed before people can turn practice with AI into lasting skill.

5The study has limits. It looked only at patent lawyers, not students, and the group of junior lawyers was small. Google paid for the research, and the team could not fully check that no one used AI on the final task.

Words　単語

professionalsA2名専門家、プロ（professional の複数形）people who do a skilled job for money
peer-reviewed外形（論文が）ほかの研究者の審査を受けたchecked by other experts before it is printed
randomly外副無作為に、でたらめにby chance, without a plan or order
assistantA2名助手、アシスタントa person or thing that helps with work
draftsB2名下書き（draft の複数形）early forms of a piece of writing
overallB2副全体としてwhen you look at everything together
advantageA2名有利な点、優位something that helps you do better than others
expertise外名専門知識、専門的な技術special skill or knowledge in one area
lastingB1形長続きするstaying for a long time
concludedB1動結論づけた（conclude の過去形）decided what is true after thinking about facts

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

  1. What kind of workers did the study follow?

    A. High school teachersB. Patent lawyersC. Computer engineersD. University students

  2. How were the lawyers put into the AI group?

    A. They asked to join it.B. Their bosses chose them.C. They were picked randomly.D. Only the youngest joined it.

  3. In the final task without AI, where did the advantage of the AI group come from?

    A. Entirely from senior lawyersB. Entirely from junior lawyersC. Equally from all lawyersD. From lawyers who never used AI

  4. How did the junior AI users' scores change compared with other juniors?

    A. All of them got top scores.B. All of them got poor scores.C. They stayed exactly the same.D. Fewer got middle scores, and more got poor or good scores.

  5. Which is one limit of the study?

    A. It lasted ten years.B. It did not include any AI.C. It looked only at patent lawyers.D. It was paid for by students.

2. True or False　正しければ T、違っていれば F

  1. (     ) The study has already been peer-reviewed.

  2. (     ) The experts who scored the work did not know which group each lawyer was in.

  3. (     ) The junior lawyers had been working for less than seven years.

  4. (     ) The team was sure that no one used AI on the final task.

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）

randomlyexpertisepeer-reviewedoveralldraftsprofessionals

  1. The authors concluded that basic _____ may be needed before people can turn practice with AI into lasting skill.

  2. Experienced _____ who worked with AI for three months did better on a later task without AI, but beginners did not gain on average, according to a new study.

  3. While they had the tool, the AI users wrote better patent _____ than the other group.

  4. The study was posted as a working paper by the National Bureau of Economic Research and has not yet been _____.

  5. Lawyers who had used the tool did better _____, but the entire advantage came from senior lawyers.

  6. About two-thirds of the lawyers were _____ given an AI writing assistant built by Google, while the rest worked without one.

4. Word Order　語を並べかえて正しい文にしよう

  1. ( the / lawyers / new / inventions / Patent / that / protect / write / documents ) .

  2. ( also / out / became / results / The / lawyers' / spread / more / junior ) .

  3. ( limits / The / study / has ) .

5. Think and Write　英語で答えてみよう

  1. Should students be allowed to use AI for homework? Give reasons for your opinion.（生徒が宿題にAIを使うことは認められるべきだと思いますか？理由をつけて意見を述べてください。）

  2. Based on this study, when do you think it is better to start using AI to learn something new?（この研究をふまえると、新しいことを学ぶのにAIを使い始めるのは、いつがよいと思いますか？）

Answers　解答

1. (1) B　(2) C　(3) A　(4) D　(5) C

2. (1) F　(2) T　(3) T　(4) F

3. (1) expertise　(2) professionals　(3) drafts　(4) peer-reviewed　(5) overall　(6) randomly

4.

  1. Patent lawyers write the documents that protect new inventions.
  2. The junior lawyers' results also became more spread out.
  3. The study has limits.

全文訳

1新しい研究によると、3か月AIといっしょに働いた経験豊かな専門家は、そのあとのAIなしの課題でよい結果を出しましたが、初心者は平均すると伸びていませんでした。この研究は、全米経済研究所（NBER）のワーキングペーパーとして公開されたもので、まだほかの研究者による審査は受けていません。

2マサチューセッツ工科大学（MIT）のデビッド・オーターさんとグーグルのチームをふくむ研究者たちは、アメリカの11の法律事務所で働く特許の弁護士133人を追いかけました。特許の弁護士は、新しい発明を守るための書類を書く人たちです。弁護士の約3分の2がくじ引きのように無作為に選ばれ、グーグルが作ったAIの文章アシスタントを渡されました。残りの人はそれを使わずに働きました。だれがどちらのグループかを知らされていない専門の弁護士たちが、その仕事に点数をつけました。

3道具を使えるあいだ、AIを使った人は、もう一方のグループより出来のよい特許の下書きを書きました。しかし3か月後、弁護士全員が、AIを使わずに特許の下書きを見直して赤を入れるよう求められました。道具を使っていた弁護士は全体としてよい結果を出しましたが、その差はすべてベテランの弁護士によるものでした。働き始めて7年未満の若手の弁護士は、平均すると伸びが見られませんでした。

4若手の弁護士の結果は、ばらつきも大きくなりました。AIを使わなかった若手とくらべると、真ん中の点数の人が減り、低い点数の人と高い点数の人がどちらも増えました。著者たちは、AIを使った練習を長く続く力に変えるには、その前に基本の専門知識が必要なのかもしれない、と結論づけました。

5この研究には限界もあります。調べたのは特許の弁護士だけで学生ではなく、若手の弁護士の人数も少なかったのです。研究のお金を出したのはグーグルで、最後の課題でだれもAIを使わなかったかどうかを、チームは完全には確かめられませんでした。

文法ポイント：受け身＋to不定詞（be asked to ～）と現在完了の受け身（has not been＋過去分詞）

be asked to＋動詞の原形で「～するよう求められる」。本文の all of the lawyers were asked to review ... は「全員が見直すよう求められた」。また has not yet been peer-reviewed は現在完了の受け身で「まだ審査されていない」。

出典：National Bureau of Economic Research (NBER)、PsyPost 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/2026-11-09-ai-skills-experts-study/