Alphabet’s Pichai Urges ‘Proportionate’ AI Regulation

Sundar Pichai, chief executive of Google parent Alphabet, has urged regulators to take a “proportionate approach” as they look to impose rules governing artificial intelligence.

“There is no question in my mind that artificial intelligence needs to be regulated. It is too important not to,” Pichai said in remarks prepared for a conference in Brussels organised by the Bruegel think tank.  “The only question is how to approach it.”

Pichai’s talk comes as the European Commission prepares to deliver proposals on new AI regulations.

The EU plans suggest imposing a ban of up to five years on the use of automated facial recognition in public places while regulators study the technology’s possible impact on areas such as individual privacy.

‘Nefarious’ tech

Pichai acknowledged that facial recognition could be used for “nefarious reasons”, and said Google is not selling general-purpose facial recognition tools while it establishes policies and technical safeguards.

But he didn’t speak out against rivals such as Amazon that do sell the controversial technology.

Pichai said regulators should refrain from establishing blanket rules covering all AI, instead taking into account the way it’s used in different fields.

In an area such as self-driving cars “appropriate new rules” should be introduced, while in the healthcare field he said existing frameworks could be extended to cover AI-assisted tools.

“Sensible regulation must also take a proportionate approach, balancing potential harms with social opportunities,” Pichai said.

He said this was especially true in areas that are “high risk and high value”.

Standards

Pichai also called on governments to work on aligning their regulatory regimes to create global standards.

The EU’s new proposals indicate it is taking a different tack from the “light touch” guidelines published by the US government earlier this month, which focus on limiting authorities’ overreach.

In the guidelines, the US urged the EU to avoid an aggressive approach.

The EU’s proposals, on the other hand, said a strong enforcement framework involving the participation of national authorities would be necessary to ensure technologies such as automated facial recognition are not abused.

Pichai singled out “deep fakes”, the use of AI to manipulate video and audio clips, as another area of concern.

He said Google has released open datasets to help researchers build better tools to detect such materials.

“Companies such as ours cannot just build promising new technology and let market forces decide how it will be used,” Pichai said.

“It is equally incumbent on us to make sure that technology is harnessed for good and available to everyone.”