Scott Wiener heads to Congress with AI regulation high on his agenda

Exercise 1
Vocabulary
Repeat each word, its meaning and the example sentence out loud.
frontier AI
adjective/ˈfrʌntɪər eɪ.aɪ/cutting-edge artificial intelligence systems at the leading edge of capability, typically large language models or advanced models that push technical boundaries
Our frontier AI projects require compliance with stricter testing protocols under the new regulatory framework.
recursive self-improvement
noun phrase/rɪˈkɝːsɪv ˌsɛlf ɪmˈpruːvmənt/a theoretical phenomenon where an AI system enhances itself and uses those improvements to enhance itself further, potentially spiraling beyond human control
The research team flagged recursive self-improvement as the primary risk scenario we need to engineer safeguards against.
sleepwalk into
phrasal verb/ˈsliːpˌwɔːk ˈɪntu/to move passively or unaware toward a problematic situation without deliberate action or full recognition of consequences
Without proper governance structures, we risk sleepwalking into a scenario where AI development outpaces our ability to ensure safety.
nimble
adjective/ˈnɪmbəl/quick, flexible, and able to adapt swiftly to changing circumstances or requirements
We need a nimble regulatory body that can adjust requirements as AI capabilities evolve rather than rigid rules that become obsolete.
real teeth
noun phrase/ˈrɪəl tiːθ/substantive enforcement mechanisms and genuine consequences, not symbolic or toothless regulations that lack practical power to compel compliance
The new AI safety legislation needs real teeth—meaningful penalties and enforcement capability—rather than voluntary guidelines.
Exercise 2
Article
Read the article aloud on your own, or listen to it first.
Scott Wiener heads to Congress with AI regulation high on his agenda
Tap any word to see what it means.
Scott Wiener, a California state senator with a decade-long focus on AI governance, has won the congressional race for Nancy Pelosi's former San Francisco seat and intends to make federal AI regulation central to his legislative agenda.
During his tenure in California's legislature, Wiener authored SB 1047 in 2024, which would have been the nation's first comprehensive frontier AI legislation. Though California's governor Gavin Newsom vetoed it, Wiener subsequently sponsored SB 53, a more measured version that Newsom signed into law.
Both measures have already influenced proposed federal legislation.
Wiener's congressional victory comes at a moment of escalating concern within AI safety circles. Just before his electoral success was confirmed, the AI research nonprofit METR released findings that many observers interpreted as evidence that Anthropic had achieved recursive self-improvement—a scenario where AI systems contribute meaningfully to their own advancement and progress could spiral beyond human control.
Speaking shortly after his victory, Wiener emphasized the urgency of the moment. He expressed frustration that Congress has failed to act decisively before year's end and warned that the window for meaningful intervention is rapidly narrowing.
"We're in a moment that is very dangerous," he stated, echoing language typically deployed by AI safety researchers. "We need to slow this down because it's moving very quickly, and we could sleepwalk into a disaster."
Wiener's electoral campaign drew substantial financial backing from Public First Action, a nonprofit dedicated to AI regulation that receives considerable funding from Anthropic. This connection drew accusations from opponents claiming he was compromised by industry interests.
Wiener countered that such attacks were disingenuous, noting his consistent track record passing actual AI safety legislation despite industry opposition.
His policy priorities reflect core concerns within the AI safety movement. He champions a robust federal AI regulator with genuine enforcement authority, coupled with whistleblower protections, mandatory testing and auditing regimes, and a kill switch requirement enabling intervention when systems exhibit dangerous behavior.
He emphasizes that Congress must leverage its oversight powers to interrogate AI companies rigorously.
Wiener has expressed interest in serving on a prospective congressional select committee on artificial intelligence, should one be established. He welcomes policy proposals circulating among colleagues, particularly legislation from fellow Californian Ro Khanna proposing an empowered federal regulator.
As a veteran state legislator, Wiener opposes broad federal preemption of state AI laws, preferring to preserve federalism. He argues that states function as regulatory laboratories, capable of nimble adaptation in ways Congress struggles to achieve.
Acknowledging the landscape, Wiener remains only cautiously optimistic about congressional capacity for meaningful AI regulation, citing the legislature's historical inability to constrain the technology industry effectively. Moreover, he expressed disquiet about the incoming Trump administration's dismissive stance toward catastrophic AI risk and its reluctance to pursue international collaboration—which Wiener deems essential for preventing worst-case scenarios.
Yet Wiener discerns an opening in presidential unpredictability. He expressed guarded hope that while troubling signals have emerged, catastrophic outcomes have not yet materialized.
He remains uncertain whether a genuine crisis might compel presidential action, though he characterizes such reliance as fraught.
Notably, Wiener distinguishes between genuine AI safety concerns and the electorate's opposition to data centers. Forty-one ballot measures opposing local data center development succeeded in recent elections, which might superficially appear to bolster Wiener's regulatory agenda.
However, he contends that most voters are driven by parochial grievances—concerns about electricity costs, water consumption, and neighborhood impacts—rather than existential AI risk assessment.
His congressional mandate positions him to influence federal technology policy at a pivotal moment, though formidable obstacles persist in translating his legislative ambitions into enacted law.
Discuss this story out loud with our AI speaking coach.
It asks one question at a time, at your level, and gives you feedback at the end.
Start the discussion Retell it3,000+ learners
Also works inChatGPTandClaude
Exercise 3
Discussion
Have a discussion based on the following questions.
- 1.Wiener positions himself as prioritizing AI safety over industry interests, yet his campaign received substantial backing from an Anthropic-affiliated nonprofit. How would you navigate such apparent tensions between funding sources and stated policy commitments in your own organization?
- 2.Wiener advocates for a strong federal regulator with enforcement teeth, but acknowledges Congress's historical inability to constrain technology companies effectively. What structural or political barriers do you perceive that might prevent Congress from acting decisively on AI regulation?
- 3.Wiener opposes federal preemption of state AI laws, arguing states should function as regulatory laboratories. What advantages and disadvantages do you foresee in allowing fragmented, state-level AI governance versus unified federal standards for your organization's compliance obligations?
- 4.Wiener emphasizes existential risks from recursive self-improvement, but acknowledges most voters prioritize local concerns like data center impacts. Where should policymakers focus regulatory efforts—on speculative catastrophic scenarios or on addressing tangible harms voters experience today?
- 5.The incoming administration views catastrophic AI risk as a hoax and resists international collaboration. How might a geopolitically fragmented approach to AI governance affect your team's ability to develop and deploy AI systems globally?
Exercise 4
Further discussion
Have a discussion based on the following questions.
- 1.What mechanisms could Congress implement to maintain regulatory nimbleness as AI capabilities evolve rapidly, preventing regulation from becoming obsolete faster than it can be updated?
- 2.How should policymakers balance fostering innovation and economic competitiveness against imposing safety requirements that might slow deployment or increase development costs, particularly given international competition from jurisdictions with lighter regulatory frameworks?
- 3.If recursive self-improvement and genuine loss of control of AI systems become demonstrable rather than theoretical risks, what emergency regulatory or technical measures should societies have pre-positioned to respond rapidly?
- 4.How can regulatory bodies and technologists establish credible oversight mechanisms for frontier AI development when the underlying science remains contested and expertise is concentrated within industry?
Discuss this story out loud with our AI speaking coach.
It asks one question at a time, at your level, and gives you feedback at the end.
Start the discussion Retell it3,000+ learners
Also works inChatGPTandClaude
Source
This lesson is our own retelling, for English learners, of a story from MIT Technology Review.
Related stories

A Mathematician and an AI System Both Claim to Have Solved the Riemann Hypothesis
22 November 2026 · C2

Anthropic's IPO: Can a safety-focused AI company survive shareholder pressure?
22 November 2026 · C1

Self-Teaching AI Systems: A Breakthrough or a Risk?
22 November 2026 · B2

AI and sensors help warn communities about dangerous landslides
22 November 2026 · B1



