Palantir technology chief Shyam Sankar has argued that AI companies must take responsibility for engineering safe products. In an October 11 interview, he proposed regulating applications through existing institutions and liability rules. He also said the industry should demonstrate benefits for jobs and prosperity, while facing questions about commercial incentives and the limits of human control.
Palantir, the software company, shouldered questions about AI safety when its chief technology officer Shyam Sankar appeared on Face the Nation on October 11. In the CBS News interview, he argued that technology companies must undertake the engineering work needed to make their products safe. He rejected the suggestion that he opposed every safeguard and maintained that producers could not pass their responsibility for safe design to someone else.[1]
Sankar proposed regulating the application of AI and pointed to the US Food and Drug Administration for medical uses. He argued that existing liability law offered a starting point. His car analogy distinguished a manufacturer’s responsibility for a defective product from a user’s responsibility for deliberately causing damage with it. Interviewer Margaret Brennan questioned whether losing human control would change that comparison; Sankar maintained that people retained control.[1]
Sankar said the industry also needed to demonstrate benefits that produce job growth and prosperity, beyond small conveniences such as travel planning. Brennan raised commercial incentives as a possible explanation for resistance to guardrails. She cited CBS polling in which 50 per cent anticipated more problems than solutions from AI, while 29 per cent expected more solutions and 21 per cent were unsure. Government limits on developing and using AI attracted support from 45 per cent.[1]
