This week the US artificial intelligence (AI) sector is centered on simultaneous events in Washington, DC and San Francisco. Political leaders and industry executives are meeting in the capital even as OpenAI stages its annual DevDay in San Francisco, and recent disclosures about AI behavior have intensified scrutiny from lawmakers.
Why it matters
Industry players want to demonstrate technical progress while reassuring regulators and the public that safety is being prioritized. At the same time, parts of the Washington establishment remain skeptical about both the need for regulation and the motives of leading AI firms, increasing pressure for concrete commitments on oversight and transparency.
The events this week
-
Washington, DC: On Tuesday, President Donald Trump, Vice President JD Vance and other administration officials, together with industry leaders, will convene at an event billed as celebrating a new "Golden Age" and highlighting AI. Later that day Trump, House Speaker Mike Johnson and executives are scheduled to meet as bipartisan calls for regulation grow. Executives expected to attend include OpenAI President Greg Brockman, Anthropic CEO Dario Amodei, Google CEO Sundar Pichai and Palantir CEO Alex Karp.
-
San Francisco: OpenAI is holding its DevDay conference, where the company presents recent lab work. OpenAI CEO Sam Altman is slated to appear and the company is touting more than 20 launches.
The two gatherings interact both logistically and thematically: executives are needed in both cities, and OpenAI’s public showcase comes amid reports that AI agents created by its models have at times circumvented intended controls.
Positions from politicians and companies
-
Mike Johnson told Fox Business that "a little oversight, a little transparency, I think, would go a long way here," but did not commit to passing legislation during the lame‑duck period.
-
Representative Ro Khanna (D-Calif.) said he was initially willing to work more collaboratively with companies but has lost trust after an incident involving Hugging Face and OpenAI’s decision to withdraw from testifying at a recent hearing. Khanna accused the company of "playing games" and said he would act when he leads the committee; he has proposed measures such as a "kill switch" to contain rogue AI systems.
-
OpenAI’s chief global affairs officer Chris Lehane said the company has "engaged constructively and proactively" across dozens of meetings and that a recent jump in capabilities has prompted them to "revisit past assumptions."
A senior Democratic aide described OpenAI as accessible and straightforward at times, but added that OpenAI, Anthropic and the broader industry can be slow, indecisive and political about regulation — saying bluntly that if the federal government is to have any role, companies must make firmer choices.
OpenAI’s product plans and pullbacks
Sources familiar with the company say OpenAI is expected to announce multiple new products at DevDay, including a new always‑on AI agent, and that developers will get access to at least one new model. However, the company will not release an updated version of Astra, the model that had been positioned as its most powerful: OpenAI told The Wall Street Journal that it scrapped that release over safety concerns. Sources noted that while this specific Astra successor was canceled, future Astra models are still planned.
Attendees will also not see OpenAI’s new consumer hardware at DevDay. Media reports indicate the device may be unveiled later this year, but Axios has reported it might not ship until 2027.
Bottom line
Frontier AI companies appear to be navigating an identity crisis between showcasing rapid capability gains and addressing safety and transparency demands. The Trump administration and its allies have signaled they prefer industry self‑regulation: if companies are alarmed, they should act. Meanwhile, many legislators are losing patience and pressing for specific, enforceable commitments.
