We worked with Encode AI, The Midas Project and Guidelight AI Standards to publish a report on how the public (and especially investors) should view SpaceX’s IPO in light of the company’s recent merger with xAI, Elon Musk’s AI model development company.
As our report details, xAI has ranked behind the leading frontier developers in every major published assessment of AI safety practices. This nonchalance towards safety is material to investors because, as we’ve seen, gaps in safety practices lead to increased regulatory burdens, litigation exposure, and public crises. Some recent press suggests that “SpaceXAI” (as they are calling the xAI business within SpaceX) may pivot from model developer to AI infrastructure provider, and it’s possible it will take safety more seriously than pre-SpaceX xAI. But if SpaceXAI continues model development with the same lax approach to safety as xAI, the investing public should be aware of the risks that creates for SpaceX and how those risks will evolve and magnify as models become more capable and are deployed more widely in society.
With SpaceX’s IPO on the horizon, investors should take action to ensure they have all the facts, asking about (i) xAI’s track record of safety incidents and the resulting regulatory and litigation exposure, (ii) SpaceX’s forward-looking expectations for its AI models’ capabilities and the new risks such capabilities might create, and (iii) how the company intends to ensure its products remain safe as those new risks emerge.
Read the full report at spacexai-risks.org.