SUMMARY

On June 16, the Trump administration forced Anthropic, one of the country’s leading artificial intelligence companies, to disable its new and powerful AI model, citing security concerns. It’s an unprecedented move that comes as the government struggles to define the guardrails for an industry whose technology is rapidly advancing.

View the transcript of the story.

NOTE: If you are short on time, watch the video and complete this See, Think, Wonder activity: What did you notice? What did the story make you think about? What would you want to learn more about?

News alternative: Check our recent segments from the News Hour, and choose the story you’re most interested in watching. You can make a Google doc copy of discussion questions that work for any of the stories here.

WARM-UP QUESTIONS

1. What is Anthropic, and what does it produce?
2. Where in the U.S. government (name of department) are federal workers customers of Anthropic?
3. What specific models from Anthropic have been vulnerable?
4. When did the U.S. government decide to change their stance on AI regulation, according to this piece? (provide more than one date, if necessary)
5. How do the steps taken by the U.S. government against Anthropic compare to their policies in the past regarding AI regulation?

ESSENTIAL QUESTIONS

1. Amazon CEO Alex Jassy raised concerns about the guardrails around Anthropic’s models, saying that researchers had found a way to get around them. What do you think this means for future AI model security and development?
2. Anthropic has argued that other companies have similar issues but haven’t faced the same restrictions. Why has the government’s approach to Anthropic been different, according to this piece? Do you agree with this approach? Explain.

Media literacy: Read over the statistic provided by IBM at the end of the video. After, answer the following question:

Consider how AI tools are being used in the government and public life. What does it mean that it takes very little time on average to hack into these AI models? What are the risks of continued use of AI without restrictions or cybersecurity fixes?

WHAT STUDENTS CAN DO

Read over this article on the most recent executive order on AI, which breaks down what the order means for how AI companies can approach innovation and potential regulation of powerful models. After students are finished reading, they should discuss the potential benefits and concerns of using AI if major risks are still present. Some guiding questions can be:

• What does it mean that the U.S. government is potentially both a ‘customer’ and a ‘shareholder’ of Anthropic?
• Does too much involvement in a company with security issues risk key U.S government systems being vulnerable to hacking attempts?

Written by Alyssa Tinoco, News Hour Classroom intern, and PBS News Hour's Vic Pasquantonio

Sign up to receive our weekly newsletter with Daily News Lessons and community events.

To provide feedback on News Hour Classroom's resources, including this lesson, click here.