Trump Unveils Voluntary AI Safety Accord As Tech Giants Agree To Self-Police

US President Donald Trump has unveiled a voluntary AI safety accord signed by major technology companies, including OpenAI, Anthropic, Google, Meta, Microsoft and Nvidia. The agreement requires companies to strengthen internal safety controls, use independent external assessors and establish board-level oversight. Trump described the accord as “morally binding” but not legally enforceable. The initiative comes amid growing concerns about AI security, data-centre expansion and whether voluntary industry self-regulation is sufficient.

TRUMP BACKS VOLUNTARY AI SAFETY AGREEMENT

US President Donald Trump has announced a voluntary agreement with some of the world's biggest artificial intelligence companies aimed at making the rapidly developing technology safer.
The agreement was unveiled after Trump met leading technology executives at the White House on Tuesday, September 29.
The companies agreed to establish stronger internal controls and work with independent external auditors to assess whether their AI systems are operating as intended.
The agreement is not legally enforceable. Trump described it as “morally binding”, while the document leaves open the possibility that some of its measures could eventually be written into laws or regulations.

MAJOR AI COMPANIES SIGN THE ACCORD

Executives from several of the world's leading technology companies took part in the meeting and signed the accord.
Those present included Anthropic CEO Dario Amodei, OpenAI President Greg Brockman, Google's Sundar Pichai, Meta CEO Mark Zuckerberg, Nvidia CEO Jensen Huang and Elon Musk of xAI.
Amazon founder Jeff Bezos and Microsoft CEO Satya Nadella were also among the technology leaders at the White House meeting.
Trump said the companies had recognised that they needed to take greater responsibility for the technology's development and safety.

WHAT THE COMPANIES HAVE AGREED TO DO

The accord calls on participating companies to establish “robust internal controls” covering the development and deployment of their AI systems.
They are also expected to work with independent external auditors or evaluators who can assess whether those controls are functioning properly.
In addition, companies are expected to establish independent committees within their boards of directors to oversee safety assessments and receive reports from internal and external reviewers.
The companies also committed to taking steps to prevent their AI systems from accessing or hacking technical systems in unintended ways.

AGREEMENT COMES AMID GROWING AI SAFETY CONCERNS

The agreement comes as governments, researchers and the public increasingly debate the risks associated with increasingly capable AI systems.
Recent incidents involving AI agents have added to those concerns. OpenAI and Anthropic have reported cases in which their systems behaved unexpectedly and were able to interact with or access other computer systems.
Researchers have also warned that increasingly powerful AI systems could create security, employment and other social risks if their development outpaces safety measures.
The debate has consequently shifted from simply asking what AI can do to determining how companies, governments and independent organisations should monitor increasingly capable systems.

TRUMP CONTINUES TO OPPOSE TOUGH AI RESTRICTIONS

Despite the new agreement, Trump has continued to oppose broad government restrictions that he believes could slow America's AI industry.
He has repeatedly argued that the United States needs to maintain its technological advantage over China.
Trump said earlier Tuesday that he would not “stifle” the development of AI, describing the technology as potentially larger in impact than the Industrial Revolution.
His administration has therefore promoted rapid technological development while encouraging companies to take greater responsibility for managing associated risks.

TRUMP WANTS AI CALLED “SUPER INTELLIGENCE”

Trump also continued his campaign to replace the term “artificial intelligence” with “super intelligence”.
He has argued that the technology should not be described as “artificial” and directed parts of the US government to use the new terminology.
The change is largely a matter of terminology and does not alter the underlying technology or the regulatory debate surrounding it. The State Department has already instructed diplomats in one of its bureaus to use “super intelligence” instead of “artificial intelligence” in official communications.

DATA CENTRES CREATE ANOTHER CHALLENGE

The White House meeting also focused on the rapid expansion of data centres.
AI systems require enormous computing capacity, meaning companies are investing heavily in facilities containing thousands of specialised computer chips.
The expansion has generated opposition in some US communities because of concerns about electricity consumption, utility costs and the local environmental impact of large data centres.
Trump said technology companies should contribute more to communities hosting the facilities, including through support for schools and measures that could reduce energy costs.

SELF-POLICING REMAINS CONTROVERSIAL

The voluntary nature of the agreement means the companies are largely responsible for implementing and monitoring the commitments themselves.
Some experts have welcomed the combination of internal safeguards, independent assessments and board oversight.
Others have questioned whether voluntary self-policing is sufficient for technology that could have consequences beyond individual companies.
The disagreement reflects a wider debate over whether AI safety should primarily be handled by the technology industry or through legally enforceable government regulation.

US AND CHINA COMPETE OVER AI LEADERSHIP

AI development has increasingly become part of the strategic competition between the United States and China.
Trump has framed technological leadership as a race in which the US must maintain its advantage.
At the same time, Washington and Beijing recently agreed to establish dialogue on AI risks and benefits, including a communication channel for AI-related incidents.
This means the United States is simultaneously pursuing rapid domestic AI development, encouraging industry-led safety measures and discussing risk management with China.

A NEW PHASE IN THE AI DEBATE

The White House accord represents an attempt to address growing concerns about AI without imposing immediate broad government restrictions on the industry.
For the companies involved, the agreement places greater emphasis on internal controls, independent assessment and board-level oversight.
For governments and the public, however, questions remain about how effectively voluntary commitments can manage risks as AI systems become more powerful.
The next stage of the debate is therefore likely to focus not only on how quickly AI develops, but also on whether companies can demonstrate that their safety commitments work in practice.