California's War On 'Bots' Could Be A Steep Uphill Climb
from the disinformation-nation dept
While California's new net neutrality law grabbed the lion's share of press headlines, the state last week also passed legislation attempting to fix an equally complicated problem: bots. The rise of bots has played a major role in not only helping companies and politicians covertly advertise products or positions, but they've also played a starring role in Russia's disinformation efforts. That in turn has fueled what's often not-entirely-productive paranoia, as users online accuse those they disagree with of being bots, instead of, say, just genuinely terrible but real human beings.
Last Sunday, California Governor Jerry Brown signed SB1001, a bill that requires companies or individuals to label any bots clearly as bots. The bill explains the legislation this way:
"This bill would, with certain exceptions, make it unlawful for any person to use a bot to communicate or interact with another person in California online with the intent to mislead the other person about its artificial identity for the purpose of knowingly deceiving the person about the content of the communication in order to incentivize a purchase or sale of goods or services in a commercial transaction or to influence a vote in an election. The bill would define various terms for these purposes. The bill would make these provisions operative on July 1, 2019."
While this is well intentioned, you can see how enforcement could prove to be a bit of a problem. Most of the scientific analysis of what constitutes a bot hasn't been particularly solid. Hamilton 68, for example, widely cited in the media as a definitive and scientific way to measure Russian bot influence online, has been widely derided as little more than educated guesswork at best.
That's not to say that "bots" aren't a problem. A disclosure from Twitter back in June noted that the company had "identified and challenged" more than 9.9 million accounts each week last May for being "potentially spammy or automated accounts." Those numbers were up dramatically from the 6.4 million per week in December 2017, and the 3.2 million million per week in September. Initially, fear of hampering ad impression metrics likely hindered many companies' interest in addressing this problem, but the exposure of Russia's online gamesmanship seems to have shifted that thinking dramatically.
California's efforts on this front were born from genuine annoyance at the problem. State Senator Robert M. Hertzberg, the bill's author, ran into the problem face first when numerous fake Facebook and Twitter accounts began berating him for a bail reform bill he proposed earlier this year. In a report in the New York Times last July, he found himself inundated with bogus attacks by bots, who did everything they could do to scuttle his proposal (note how he didn't clarify how he differentiated between bots or just terrible people).
I spoke briefly about California's new legislation with the EFF's Jeremy Gillula, who was quick to highlight the enforcement and free speech problems inherent in California's attempt to fix a problem we don't fully understand:
"Enforcement is definitely going to be difficult. Governments aren't in a good position to be able to tell whether or not an account on social media is a bot or not. And we wouldn't want them to be--that would require that they know or are able to easily find out where, when, and how every post to social media was made. That would destroy any possibility of anonymous speech online."
Gillula told me that the original bill proposed by Hertzberg was much worse, in that it would have made chatbots like Olivia Taters illegal simply because it's not clearly labeled as a bot. The EFF also successfully encouraged Hertzberg's office to eliminate a proposal that would have forced all websites to have functions letting users report bots. Again, because of the problems stated above (a culture of paranoia and difficulty differentiating bots from ordinary human jackasses), the end result could have been potentially catastrophic for speech online:
"We've seen this sort of takedown/labeling regime abused time and again in the past. For example, Twitter and Facebook users can already report posts they think violate the platforms’ community standards. Online platforms also have to deal with complaints under the Digital Millennium Copyright Act, or DMCA, a federal law meant to protect copyright online, which forces platforms to take down content. In both cases, malicious users have figured out how to abuse these reporting systems so that they can get speech they don’t like erased from the Internet. The targets victims of such abuse have been Muslim civil rights leaders, pro-democracy activists in Vietnam, pro-science accounts targeted by anti-vaccination groups, and Black Lives Matter activists, whose posts have been censored due to efforts by white supremacists."
One one hand, you can see how bringing a little more transparency to who is running bots and what their motivations are could be helpful. But actually identifying bots and removing them is currently more art than science. In the interim, it's probably a good idea to avoid creating solutions that, in time, could create even bigger problems (like an entirely new takedown system open to abuse) than the one you're trying to fix.
Thank you for reading this Techdirt post. With so many things competing for everyone’s attention these days, we really appreciate you giving us your time. We work hard every day to put quality content out there for our community.
Techdirt is one of the few remaining truly independent media outlets. We do not have a giant corporation behind us, and we rely heavily on our community to support us, in an age when advertisers are increasingly uninterested in sponsoring small, independent sites — especially a site like ours that is unwilling to pull punches in its reporting and analysis.
While other websites have resorted to paywalls, registration requirements, and increasingly annoying/intrusive advertising, we have always kept Techdirt open and available to anyone. But in order to continue doing so, we need your support. We offer a variety of ways for our readers to support us, from direct donations to special subscriptions and cool merchandise — and every little bit helps. Thank you.
–The Techdirt Team
Filed Under: bots, california, labelling, misrepresentation
Reader Comments
Subscribe: RSS
View by: Time | Thread
The odds are...
[ link to this | view in chronology ]
[ link to this | view in chronology ]
People who act like bots.
Thanks for calling Central Services. I'm sorry. Due to staff shortages, Central Services cannot take service calls between 2300 and 0900 hours.Have a nice day. This has not been a recording.
[ link to this | view in chronology ]
Re:
[ link to this | view in chronology ]
Obligatory
[ link to this | view in chronology ]
Apparently the all seeing benevolent overlords want everyone to properly identify themselves on the web, why should bots be any different?
[ link to this | view in chronology ]
Re:
[ link to this | view in chronology ]
Re:
Nobody ever obeyed those laws except as a prank.
[ link to this | view in chronology ]
Should be easy
If a bot is discovered faking it, the account is deleted.
I love the idea. Let's do it.
[ link to this | view in chronology ]
Re: Should be easy
This sort of law reeks of politicians panicking because they can no longer control public debate, and will have all sorts of unintended consequences that fly under the radar because it does small amounts of damage to an individual but a large amount of damage to society when all those little bits are added up. Also there is the risk of this being abused by those who would stamp out opinions and activities that they disagree with.
[ link to this | view in chronology ]
Re: Re: Should be easy
Further, you'd assume even if your elaborate "bot" that does twitter things for you is very active, it's not posting for you is it? Then it would be pretending to be a human.
I'm not sure what's hard to understand here. You aren't going to be able to fabricate a gray area here. It's very cut and dry.
[ link to this | view in chronology ]
Re: Re: Re: Should be easy
You claim that an automated process that sends out notifications on my social media accounts in response to my own actions (new blog/podcast post, for instance) should require my account be labeled a bot because despite posting the truthful information I want posted but otherwise would need to manually fat finger in and the accounts being otherwise used for content I personally produce, all because I automate non social media content notifications?
Its true that the process is 'pretending to be me', but that's because I tell it to post these things. And it is otherwise me posting. Are you somehow harmed by the knowledge that my blog post tweet is somehow artificial and not homegrown link copy-pasting?
Your bot example is clearly a bot, just not a content bot. Since it doesn't post anything, it clearly wouldn't apply to this situation at all, but that adds nothing to your case. That complicated automated process is still a 'bot'.
[ link to this | view in chronology ]
Re: Re: Re: Should be easy
It seems to me *you* are a bot...posting nonsense, missing the point of the article, claiming "it's easy".
In your reply, you will be screaming to the contrary...just like a bot. You probably post from a VPN address...just like a bot.
Hint: It can be very hard to identify bots, especially the malicious ones, and it will become an on-going cat-and-mouse game. And that's *before* the lines get blurred, like when I write my Farcebook posts with a tweet at the top, and that gets auto-tweeted. That human or a bot there??? Or are you one of 30 sockpuppets of just one GRU agent??
[ link to this | view in chronology ]
Re: Re: Re: Should be easy
Also, many an IRC channel runs a bot that catches and responds to FAQ's. What about people using NightBot on YouTube and twitch chat on their channels, is it name enough identification.
Like many things, the benign uses fly under the political radar, and are never considered when a bad, according to politicians, use surfaces.
This issue is nowhere near all black as you seem to think it is, but includes all shade from white to black.
[ link to this | view in chronology ]
Re: Should be easy
Maybe a button for reporting suspected bots to the proper authorities?
[ link to this | view in chronology ]
Re: Re: Should be easy
[ link to this | view in chronology ]
Re: Re: Re: Should be easy
My phone is part of the conspiracy, obviously it wanted to change Turing to Touring.
He - my phone actually IS a bot. Siri and Autocorrect are both bots according to this California law!
[ link to this | view in chronology ]
Re: Re: Re: Re: Should be easy
[ link to this | view in chronology ]
This explains a lot
[ link to this | view in chronology ]
Re: This explains a lot
If he actually was replaced with a bot, one programmed to endlessly spew out random, incoherent and nonsensical sentence-fragments copy/pasted from alt-right Twitter accounts, how would we tell the difference? He more or less does that already. If anything, he'd presumably come across as slightly more intelligible and less dangerously chaotic than usual.
More pertinently, if some enterprising, shadowy party were to do this in real life, who exactly would it inconvenience? His removal and replacement would hardly seem to constitute a catastrophic loss for world democracy.
The more I think about it, the more this sounds like a plan.
I wonder if we can get this crowdfunded?
--dg100
[ link to this | view in chronology ]
Re: This explains a lot
[ link to this | view in chronology ]
Re: Re: This explains a lot
I happen to know a few lizardfolk who would make vastly better Presidents than the current inhabitant of the Oval Office, so I think your comparison falls flat.
[ link to this | view in chronology ]
Just another attempt to suppress speech someone doesn't like.
[ link to this | view in chronology ]
Re:
[ link to this | view in chronology ]
Inb4 Ajit Pai & Co sue CA over this law too. Otherwise, it'll be a little tougher for him to get as much "public support" for his future actions.
[ link to this | view in chronology ]
I believe their system for deciding who a bot is and isn't is quite simple.
Anyone who disagrees with them is a bot and anyone who agrees with what they say isn't a bot.
[ link to this | view in chronology ]
Terrible People
> inundated with bogus attacks by bots, who did everything
> they could do to scuttle his proposal (note how he didn't
> clarify how he differentiated between bots or just
> terrible people).
So the only options here are bots and terrible people?
If you opposed the idiotic idea of letting people arrested for serious crimes out of jail until the justice system ground it way around to their trial, you're 'just a terrible person'?
Or does this apply to whatever Hertzberg proposes in general? Any non-bot who takes issue with him is 'just a terrible person'?
GTFO
[ link to this | view in chronology ]