Microsoft cut ties with some customers over their plans for AI
Microsoft puts AI business cases to ethical committee before committing, claims the company's lead AI researcher
Microsoft has severed links with a number of customers and potential customers over concerns over the way in which they were planning to use the company's nascent artificial intelligence technology.
Eric Horvitz, a technical fellow and director of Microsoft Research Labs, is becoming increasingly worried that some companies may exploit AI technology in the wrong way, according to Geekwire.
Speaking at Carnegie Mellon University & K&L Gates Conference on Ethics and AI yesterday, Horvitz explained that the software giant is not afraid to take action against developers of rogue AI.
He detailed Microsoft's artificial intelligence and machine learning ethics committee, called Aether (AI and Ethics in Engineering and Research).
In a keynote speech, Horvitz described the committee as having "teeth" and making an "intensive effort" to ensure that AI systems do not get out of control.
In the future, Horvitz believes that AI technologies will be able to keep themselves in check
The committee is tasked with reviewing ways in which Microsoft's AI products can be deployed by its customers, said Horvitz. Decisions even come from the senior leadership.
Horvitz said that "significant sales have been cut off" as a result of the committee intervening with customer AI plans, but he did not name firms or products.
"And in other sales, various specific limitations were written down in terms of usage, including ‘may not use data-driven pattern recognition for use in face recognition or predictions of this type," he said.
Microsoft offers a range of AI products, including the cloud-based Microsoft Cognitive Services portfolio. The latter includes facial recognition and emotion-recognition solutions.
Governments and major organisations across the world have already bought into this suite of products, although Horvitz reiterated the fact that Microsoft regulates the usage of these systems.
In the future, Horvitz believes that AI technologies will be able to keep themselves in check. "You can imagine giving systems the ability to monitor their performance," he added.
But for this technology to succeed, Horvitz said that humans and AI will need to work as a team. "We're talking about AI designs for complementarity," he concluded.
In a statement to V3, the company claimed that Horvitz was referring more to the company vetting - and rejecting - potential AI tie-ups via the Aether committee, rather than withdrawing from existing deals.
It said: "We believe it is very important to develop and deploy AI in a responsible, trusted and ethical manner.
"Microsoft created the Aether committee to identify, study and recommend policies, procedures, and best practices on questions, challenges, and opportunities coming to the fore on influences of AI on people and society."