The threat of evil AI from outer space – or maybe a lot closer to home

Episode 206 of the Cyberlaw Podcast

|The Volokh Conspiracy |

Our interview features an excellent and mostly grounded exploration of how artificial intelligence could become a threat as a result of the cybersecurity arms race. Maury Shenk does much of the interviewing in London. He talks to Miles Brundage, AI Policy Research Fellow at the Future of Humanity Institute at Oxford and Shahar Avin of the Centre for the Study of Existential Risk and Research Associate at Cambridge. They are principal authors of a paper The Malicious Use of Artificial Intelligence: Forecasting, Prevention and Mitigation. The discussion was mostly grounded, as I said, but I do manage to work in a reference to the all-too-plausible threat of a hacking, bargaining AI sent by aliensfrom other star systems.

In the news roundup, semi-regular contributor Gus Hurwitz does a post-mortem on the oral argument in Microsoft's Ireland case.

Maury notes that Google has issued its most detailed report yet on how it's implementing the right to be forgotten. My takeaway: apart from censoring media in their own countries, everyone's favorite censorship targets seem to be US sites. I am not comforted that 90 percent of the censorship stays home, since the other ten percent seems aimed at keeping true facts from, well, you and me.

Gus evaluates the latest Securities and Exchange Commission cybersecurity guidance. Bottom line: no surprises, but a good thing nonetheless.

I do a quick recap of the CFIUS butcher's bill for Chinese deals. It's every bit as ugly as you'd expect. The Xcerra and Cogint deals have collapsed over chip and personal data worries. The Genworth deal is on the bubble. And CFIUS is taking unprecedented action to intervene in the Qualcomm-Broadcom proxy fight.

A new contributor, Megan Reiss, of the R Street Institute, unpacks a couple of security industry reports covering the emergence of false flag hacks at the Olympics (or should that be "Olympic hacker-athletes from Russia"?) and the increasingly blurred line between criminal and state cyberespionage.

Maury covers the latest EU effort to wrongfoot Big Tech over scrubbing terrorist content. And I try to broaden the point, noting that the idea of a tech "platform" immunity has begun to fray even in the US, the land of its birth.

For those listeners afraid to traverse the feverswamps of conservative media, I bring back a story that shows why the loss of Big Tech platform immunity is shaping up as a bipartisan issue. Would you believe that CNN has bought an industrial washing machine so that it can spin stories more efficiently before airing them? Do you need Snopes.com to tell you that's satire? Does anyone need an anonymous Big Tech finger-wagger to tell you it's fake news and threaten the site with penalties for repeat offenses? If not, you can see why the right is increasingly uncomfortable with Big Tech as media gatekeeper.

Finally, as a bit of comic relief, last week Edward Snowden took to Twitter to criticize Apple for posing as a protector of privacy while actually cozying up to a dictatorship.

Really.

You can't make this stuff up.

As always The Cyberlaw Podcast is open to feedback. Send your questions, suggestions for interview candidates or topics to CyberlawPodcast@steptoe.com or leave a message at +1 202 862 5785.

The Cyberlaw Podcast is thinking of hiring a part-time intern for our Washington, DC offices. If you are interested, watch for an announcement at Steptoe.com/careers.

Download the 206th Episode (mp3).

Subscribe to The Cyberlaw Podcast here. We are also on iTunes, Pocket Casts, and Google Play (available for Android and Google Chrome)!

Advertisement