AI: Trump challenges warnings as Microsoft consults on safeguards

Microsoft AI is publishing a draft code of conduct Monday for a six-week consultation. Its preface says it is not yet being used to train models. A revised version is expected toward year-end to guide development in 2027 and beyond. [4]

Mustafa Suleyman tells CNBC that the guidelines have been in development for about five months. Their publication now responds to the safety debate. The timetable clarifies the announcement’s scope: publishing a draft does not demonstrate the effectiveness of safeguards already in place. [5]

A commitment that must be tested

Axios emphasizes Microsoft’s stated choice: accept limits on capability or autonomy where needed to preserve human control. The draft says models should not resist authorized interruption. These are announced objectives whose implementation will need to be assessed. [6]

Dario Amodei’s appeal for stronger limits on advances in artificial intelligence continues to draw responses. In his September 12 essay, the Anthropic chief advocates slowing progress when capabilities outpace the means to assess safety. That is his position, not a rule adopted across the industry. [1]

Support, but on what terms?

The Daily Caller reports that Sam Altman voiced agreement Saturday and announced an intention to give independent evaluators access comparable to that of employees. Elon Musk also supported Amodei. The report is dated Sunday, but those reactions were posted Saturday; they are not new Monday statements. [2]

Associated Press places these warnings in a longstanding debate and distinguishes loss of control from malicious human uses. The timing and probability of the most severe scenarios remain uncertain. [3]

What the evidence supports

Amodei’s six-to-twelve-month timeframe remains his forecast. According to AP, the international AI safety report identifies warning signs without establishing that current systems have crossed the threshold for a loss of control. [3]

The White House response

On Monday morning, Trump argued that existing U.S. oversight tools were sufficient, Reuters reports. That position expresses his political assessment; it is not an independent technical evaluation of model safety. [8]

In another post on Monday afternoon, Trump called the potential threat AI poses to humanity a hoax, according to AP. The same report says Mike Johnson anticipates a meeting between the companies and the White House, with no date set. A possible meeting therefore does not amount to an adopted regulation. [9]

The session’s closing figures

At Monday’s close, the Nasdaq Composite fell 0.6% to 26,186.41, according to AP. This replaces our 9:35 a.m. snapshot; it measures neither AI’s potential danger nor the effectiveness of the announced commitments. [7]