AI Ethics And AI Law Deeply Disturbed By AI Used To Fraudulently Influence Or Corrupt Elections

Elections are vital to democracy.

Recall the famous words of John Quincy Adams, the venerated diplomat and 6th president of the United States (1825 to 1829), whereby he stated eloquently: “Always vote for principle, though you may vote alone, and you may cherish the sweetest reflection that you vote is never lost.”

Voting is essential and to be treasured.

When it comes to elections, we aim to have voters that are able to vote without abject coercion as to what choices they make as voters. They are to vote as their free mind so decides. Your vote is your vote, and your vote alone.

Furthermore, the act of voting and the voting process is intended to suitably account for the actual votes of voters. If the voting system doesn’t accurately record and tally the true votes of the voters, we no longer have a means of knowing the will of the people. We also would indubitably doubt the outcomes of the voting process and all told suffer a weakening of a strident belief in our democracy.

Per the pointed and disturbing remark of Joseph Stalin: “The people who cast the votes don’t decide an election, the people who count the votes do.”

In recap, we have those that vote, namely the voters, and we must have a valid voting system that tallies the votes, aligned in such a fashion as to orderly undertake free and fair elections. If somehow the voters are corrupted then we have a dangerous problem that undermines our democracy. If somehow the voting system is corrupted then we also have an endangering problem that undercuts our democracy.

That all seems perhaps rather obvious.

We are daily deluged with polarizing comments and screes about how voters are being presumably “corrupted” or that the voting system is suspected of being corrupted. All manner of both reasoned and unreasoned accusations abound about voting-related fraud. There is certainly grand importance in trying to prevent such fraudulency that could lead to the corruption of our elections. No doubt about that.

One angle that is not as yet especially noteworthy but that will gradually increase in visibility and pervasiveness consists of a corrupting influence that few are particularly aware of today.

Are you ready?

Artificial Intelligence (AI).

Yes, AI can be used to undercut voters and likewise undercut our voting systems. The advent of widespread computing such as our smartphones and the Internet provides a vehicle for using AI to bend or potentially snap our election facets. We need to be cognizant that as AI continues to be crafted and adopted into our everyday systems, this same advantageous AI can be turned toward a more ominous side of usage.

For my coverage of what is commonly referred to as dual-use AI, consisting of AI that on the one hand proffers tremendous benefits and at the same time can be shifted into so-called Dr. Evil endeavors, see the link here.

A crucial point to be made is that AI can indeed be of a sizable benefit to our elections. Voters can potentially leverage AI to aid in understanding the voting issues being faced. AI can serve as a means of examining the pros and cons of societal topics that are being voted upon. AI can be used in voting systems to try and improve the voting process, such as making the act of voting easier or encouraging people to readily cast their votes.

Score a plus one for AI in voting and elections.

Unfortunately, we also need to be on the watch for and score a minus one for AI in voting and elections.

How so?

I will momentarily walk you through a myriad of ways that AI can be used to undermine voters and voting systems. This might seem discouraging and you might be tempted to suggest that maybe we should not talk of such things. All in all, a head-in-the-sand approach to AI’s adverse consequences is not a sensible way to proceed. We are going to be better off by discussing and avidly exposing the downsides of AI in our society. By doing so, we have an enhanced chance of dealing with such AI and seek to minimize or mitigate the AI pitfalls.

AI in the voting milieu is yet another concern for those that particularly have an interest in AI Ethics and AI Law. The AI Ethics arena focuses on ethical and moral elements associated with the emergence of AI. The AI Law realm in this context tends to concentrate on the governance of AI, asking whether our existing laws are sufficiently up to the task of coping with the advent of AI and proposing new AI-specific laws when legal stipulations are newly needed. For my extensive and ongoing analyses of AI Ethics and AI Law, see the link here and the link here, just to name a few.

I think it is helpful to consider two major perspectives about AI in the elections sphere:

(1) AI that benefits elections and aids both voters and our voting systems

(2) AI that undercuts and potentially corrupts both voters and our voting systems

In today’s herein discussion, I am going to mainly assess the latter of those two categories. I’ve previously touched upon AI democracy-inducing benefits in some of my postings, so it is timely and useful to take a look at the other side of the coin.

It won’t be pretty. The democratically poisonous or election-toxic use of AI is seedy and alarmingly insidious. Nonetheless, we would be remiss to ignore or discount it. Exposing how AI can be detrimental to our voting is worthy of keen attention. AI can serve as a kind of Dr. Evil to usurp our free and fair elections, which is a concern that we need to get our heads fully wrapped around.

Let’s go ahead and unpack the adverse ways in which AI can substantively taint our elections and see what we can make of it.

I’d like to first lay some essential foundation about AI and particularly AI Ethics and AI Law, doing so to make sure that the discussion will be contextually sensible.

The Rising Awareness Of Ethical AI And Also AI Law

The recent era of AI was initially viewed as being AI For Good, meaning that we could use AI for the betterment of humanity. On the heels of AI For Good came the realization that we are also immersed in AI For Bad. This includes AI that is devised or self-altered into being discriminatory and makes computational choices imbuing undue biases. Sometimes the AI is built that way, while in other instances it veers into that untoward territory.

I want to make abundantly sure that we are on the same page about the nature of today’s AI.

There isn’t any AI today that is sentient. We don’t have this. We don’t know if sentient AI will be possible. Nobody can aptly predict whether we will attain sentient AI, nor whether sentient AI will somehow miraculously spontaneously arise in a form of computational cognitive supernova (usually referred to as the singularity, see my coverage at the link here).

The type of AI that I am focusing on consists of the non-sentient AI that we have today. If we wanted to wildly speculate about sentient AI, this discussion could go in a radically different direction. A sentient AI would supposedly be of human quality. You would need to consider that the sentient AI is the cognitive equivalent of a human. More so, since some speculate we might have super-intelligent AI, it is conceivable that such AI could end up being smarter than humans (for my exploration of super-intelligent AI as a possibility, see the coverage here).

I’d strongly suggest that we keep things down to earth and consider today’s computational non-sentient AI.

Realize that today’s AI is not able to “think” in any fashion on par with human thinking. When you interact with Alexa or Siri, the conversational capacities might seem akin to human capacities, but the reality is that it is computational and lacks human cognition. The latest era of AI has made extensive use of Machine Learning (ML) and Deep Learning (DL), which leverage computational pattern matching. This has led to AI systems that have the appearance of human-like proclivities. Meanwhile, there isn’t any AI today that has a semblance of common sense and nor has any of the cognitive wonderment of robust human thinking.

Be very careful of anthropomorphizing today’s AI.

ML/DL is a form of computational pattern matching. The usual approach is that you assemble data about a decision-making task. You feed the data into the ML/DL computer models. Those models seek to find mathematical patterns. After finding such patterns, if so found, the AI system then will use those patterns when encountering new data. Upon the presentation of new data, the patterns based on the “old” or historical data are applied to render a current decision.

I think you can guess where this is heading. If humans that have been making the patterned upon decisions have been incorporating untoward biases, the odds are that the data reflects this in subtle but significant ways. Machine Learning or Deep Learning computational pattern matching will simply try to mathematically mimic the data accordingly. There is no semblance of common sense or other sentient aspects of AI-crafted modeling per se.

Furthermore, the AI developers might not realize what is going on either. The arcane mathematics in the ML/DL might make it difficult to ferret out the now-hidden biases. You would rightfully hope and expect that the AI developers would test for the potentially buried biases, though this is trickier than it might seem. A solid chance exists that even with relatively extensive testing that there will be biases still embedded within the pattern-matching models of the ML/DL.

You could somewhat use the famous or infamous adage of garbage-in garbage-out. The thing is, this is more akin to biases-in that insidiously get infused as biases submerged within the AI. The algorithm decision-making (ADM) of AI axiomatically becomes laden with inequities.

Not good.

All of this has notably significant AI Ethics implications and offers a handy window into lessons learned (even before all the lessons happen) when it comes to trying to legislate AI.

Besides employing AI Ethics precepts in general, there is a corresponding question of whether we should have laws to govern various uses of AI. New laws are being bandied around at the federal, state, and local levels that concern the range and nature of how AI should be devised. The effort to draft and enact such laws is a gradual one. AI Ethics serves as a considered stopgap, at the very least, and will almost certainly to some degree be directly incorporated into those new laws.

Be aware that some adamantly argue that we do not need new laws that cover AI and that our existing laws are sufficient. They forewarn that if we do enact some of these AI laws, we will be killing the golden goose by clamping down on advances in AI that proffer immense societal advantages.

In prior columns, I’ve covered the various national and international efforts to craft and enact laws regulating AI, see the link here, for example. I have also covered the various AI Ethics principles and guidelines that various nations have identified and adopted, including for example the United Nations effort such as the UNESCO set of AI Ethics that nearly 200 countries adopted, see the link here.

Here’s a helpful keystone list of Ethical AI criteria or characteristics regarding AI systems that I’ve previously closely explored:

• Transparency
• Justice & Fairness
• Non-Maleficence
• Responsibility
• Privacy
• Beneficence
• Freedom & Autonomy
• Trust
• Sustainability
• Dignity
• Solidarity

Those AI Ethics principles are earnestly supposed to be utilized by AI developers, along with those that manage AI development efforts, and even those that ultimately field and perform upkeep on AI systems.

All stakeholders throughout the entire AI life cycle of development and usage are considered within the scope of abiding by the being-established norms of Ethical AI. This is an important highlight since the usual assumption is that “only coders” or those that program the AI is subject to adhering to the AI Ethics notions. As prior emphasized herein, it takes a village to devise and field AI, and for which the entire village has to be versed in and abide by AI Ethics precepts.

I also recently examined the AI Bill of Rights which is the official title of the U.S. government official document entitled “Blueprint for an AI Bill of Rights: Making Automated Systems Work for the American People” that was the result of a year-long effort by the Office of Science and Technology Policy (OSTP). The OSTP is a federal entity that serves to advise the American President and the US Executive Office on various technological, scientific, and engineering aspects of national importance. In that sense, you can say that this AI Bill of Rights is a document approved by and endorsed by the existing U.S. White House.

In the AI Bill of Rights, there are five keystone categories:

• Safe and effective systems
• Algorithmic discrimination protections
• Data privacy
• Notice and explanation
• Human alternatives, consideration, and fallback

I’ve carefully reviewed those precepts, see the link here.

Now that I’ve laid a helpful foundation on these related AI Ethics and AI Law topics, we are ready to jump into the heady topic of AI that undermines voters and voting systems.

AI In Voting Is A Rising Consideration And Possible Concern

In a voting context, we know that we need to be watchful for a slew of problematic maladies such as election fraud, election manipulation, voter fraud, vote rigging, voter suppression, vote tampering, and the like.

There are lots of conventional non-AI ways that voters can be “corrupted” and that our voting systems can be corrupted. Continual awareness of voter-related matters and voting systems issues is crucial to catching troubles before harm is done. Once the harm is done, if so occurring, we need to realize what has happened and seek to resolve the harms that ensued.

We are now embarking on adding AI into the eclectic mix.

Thus, you have all of the conventional methods of corrupting elections, and as a type of “bonus” (not a savory one), we now are plopping AI on top of that morass. If we unduly shift our attention to the AI-powered methods of corrupting elections, we could inadvertently allow the conventional non-AI approaches to flourish. Don’t do that.

We need to be eagle-eyed about both:

1) Conventional non-AI methods of corrupting elections

2) AI methods of corrupting elections

Just wanted to make sure that we do not become over-enamored with the shiny new AI ways. The battle for detecting and dealing with election corruption must be fought on both land and sea, as it were, entailing both conventional non-AI and also newer AI-infused approaches.

As mentioned earlier herein, I am not going to tackle the notion of sentient AI that might corrupt our elections. That would be quite a rabbit hole to dive into. Wild speculation would arise. For example, suppose we anoint AI with bona fide legal personhood. Does the AI get to vote? Would AI get only one vote as though on par with one human vote, or would we need to allocate some substantive number of “votes” to an AI system to aptly ensure that it got properly represented? Could the AI attempt to cheat and vote more than once or for more than whatever counts it is allowed? Etc.

On and on that merry-go-round goes.

I’m sticking with non-sentient AI in this herein discourse.

Today’s non-sentient AI does not get a vote (well, some pundits disagree, see the link here). I realize that there is all manner of speculation about non-sentient AI getting a smidgeon of legal personhood, which I’ve covered in my writings, but it is not directly in the cards right now and so I am going to keep that out of this discussion.

We are going to assume for sake of reasonableness that legitimate voters must be human.

We can also assume that voting systems are likely a mixture of humans and computer-based automation that works toward counting and reporting the votes of an election.

Consider these possibilities:

• AI is used legitimately to aid election activities of voters and voting systems
• AI is used in ethically questionable ways toward election activities of voters and voting systems
• AI is used in outright unethical ways toward election activities of voters and voting systems
• AI is used in demonstrably illegal ways toward election activities of voters and voting systems

There can be quite a blurring amidst those possibilities.

Take for example the use of AI chatbots.

Suppose that someone wishing to get elected makes use of AI-based chatbots. The chatbots are programmed to appear as though they are humans on a social media platform. Each of the AI chatbots concocts a normal-looking name and seemingly is a human in that the narratives generated by the AI have a human quality to the wordings. Assume that there are no obvious means to discern whether the narrative was written by a human, might be a mimicry of what a human wrote, or could be generated text that has programmatically been devised to appear as though human written.

Human voters trying to decide on whether to vote for the candidate or instead vote for the opponent are summarily bombarded with AI chatbot chatter. For all appearances, it seems as though thousands upon thousands of “people” are weighing in as overtly and loudly voting for the candidate while thousands upon thousands of “people” are also severely trashing the opponent.

I ask you, into which of the aforementioned buckets or possibilities does this scenario reside?

Your choices are that this is legitimate, or this is borderline ethical, fully unethical, or outrageously illegal.

Take a pick.

In general, here’s the official FBI stance on election-pertinent crimes: “Fair elections are the foundation of our democracy, and the FBI is committed to protecting the rights of all Americans to vote. The U.S. government only works when legal votes are counted and when campaigns follow the law. When the legitimacy of elections is corrupted, our democracy is threatened. While individual states run elections, the FBI plays an important role in protecting federal interests and preventing violations of your constitutional rights. An election crime is generally a federal crime if: (1) The ballot includes one or more federal candidates, (2) An election or polling place official abuses their office, (3) The conduct involves false voter registration, (4) The crime intentionally targets minority protected classes, or (5) The activity violates federal campaign finance laws” (per online on the FBI website under “Election Crimes and Security”).

You would likely have a hard time contending that the use of AI chatbots was unambiguously illegal. That being said, note that state and local laws have been attempting to encompass those kinds of AI practices. Whether they will stand up to legal scrutiny is not yet entirely clear. There are also efforts of trying to capture this in AI-related federal laws.

For the moment, assume that this specific use of those AI chatbots is not illegal (in this given context, while other contexts might differ dramatically). We then land on whether it is ethical or not. If you are arguing that it is unethical, does the practice cross over into absolutely unethical or is it more akin to a borderline ethical boundary?

The odds are that the candidate using the AI chatbots would adamantly insist that it is unquestionably a legitimate use and doesn’t even in the slightest way touch into the ethics realm. They might insist that the AI chatbots never stated they were humans. Those that saw or received the messages made that presumption. That’s on them, not on the AI chatbots.

In short, the candidate might assert that if human voters allow themselves to be tricked or deceived by some form of disinformation, so be it. That’s on the shoulders of the human voters. Actually, the candidate might further argue that it wasn’t disinformation at all. If the information conveyed about the positions of the candidates was reasonably accurate, the only “twist” was that the appearance of the communications happened lots of times. Does that constitute disinformation or is it only the contents of the messages that fall within the conception of disinformation or misinformation?

This scenario so far seems to portray voter influencing and leaves a great deal of grey area about the range of ethical or legal matters involved.

Our eager candidate decides to up the ante.

Using AI, a deepfake is devised that seems to showcase the opponent saying very dastardly things. I’m sure that you’ve been noticing how advanced AI is getting at creating simulated versions of real people. This includes the visual look and mannerisms of people such as specific celebrities or others, along with copying and being able to utter new verbalized utterances that seem to be of that person.

The deepfake is released onto social media. Humans on social media opt to spread the deepfake. Thus, when you receive the deepfake, it didn’t come from an AI chatbot (though, that certainly could be the case) and instead came from a trusted colleague or friend.

You assume that since the deepfake came from someone that you know and trust, the deepfake is presumably real. Oops, you are wrong about that.

In this example, where are we now in the AI that is used for election activities?

Legitimate or questionably ethical or unethical, entirely unethical, or clearly illegal.

Here’s next a situation that we can with some relief likely agree is totally illegal.

An AI system is devised that “pretends” to be real people. It is used to cast votes that are done via mail. The AI taps into the voter rolls and figures out which people are probably not going to vote. A paper ballot is requested on their behalf. The paper ballot is filled in via the AI and mailed to the election registrar as though the vote had been cast by the actual human.

I don’t think anyone can say that this is nothing other than an illegal election activity.

An AI is voting rather than a human voter.

Not allowed.

It could be hard to catch. Assuming that the voter checks-and-balances seem to match, the devised AI might get away with this.

If the human voter perchance does vote on their own too, we might have the chance of realizing that two votes have occurred, in which case the gig might be up. Tracing the origins of the fraudulent vote might be difficult though. There is also a possibility that both votes can be nixed and ergo the human voter is getting undermined. We can only hope that at least the chance of the AI vote doesn’t get rated as the true vote while the human cast “duplicative” vote gets dumped as fakery. That hurts as a double whammy.

Let’s ratchet things up.

Some are suggesting that we ought to purposefully devise AI as a type of helpful service for voting. The AI-voting-as-a-service would at your behest assist you in voting. No longer would you need to deal with the voting paperwork. Instead, you invoke an AI-based app that will present the voting paperwork for you online and guide you through the voting process.

The AI could be trained on your voter preferences. As such, you don’t need to look at each of the candidate statements or anything else. By pressing a button, the AI will examine the entirety of the ballot and cast votes for each candidate or proposition as you might presumably have done. Upon doing so, the ballot is now ready for you to review and sign.

You are busy, and you “believe” that the AI is likely to have done a good job of choosing choices for you, therefore you earnestly sign the ballot and approve it to be cast.

Sounds pretty nifty.

Of course, suppose that somehow the AI was secretly rigged such that it voted for particular candidates or particular propositions. You might not realize the fix is in. When it comes time to sign and approve, you hardly give glance and just assume that AI did your bidding.

Are you an innocent that was taken advantage of?

Or are you a voter that wasn’t being diligent and undercut your own voting rights?

John F. Kennedy famously stated: “The ignorance of one voter in a democracy impairs the security of all.”

Yikes, some would react, that’s completely unfair to the voter that was tricked by the AI. If the AI was cybersecurity cracked to vote in ways that the voter did not know about, this is a fraudulent act by the AI or really the maker of the AI. You can’t lay the blame on the voter. Others would insist that any voter that agrees to use an AI for voting assistance owes to their own detriment the realization that they might usurp their vote.

Shifting gears, let’s take a look at AI that undercuts voting systems.

AI could be devised to:

• Trick voting systems
• Crack into voting systems
• Alter voting systems
• Takeover control of voting systems
• Intervene in the reporting from voting systems
• Deceive the humans that are running the voting systems
• Aid humans that are trying to commit fraud in voting systems
• Other

Each of those approaches tends to combine cybersecurity capacities with the use of AI. In that sense, we generally already by and large ought to be wary of those means of committing cyber crimes associated with elections.

AI has the potential to make cyber crimes more pronounced.

One key additive property is that the malicious actor can potentially perform election fraud without being so directly involved. They can be at a further arm’s length away from culpability. Another is that the fraudster might not need to know much of anything about how to commit election corruption per se. They let the AI do all the work for them.

Conclusion

We are inevitably going to find ourselves confounded by claims that AI was used to corrupt our elections.

Even if the claims are principally unfounded, the headline-grabbing potential of saying that AI was involved will be enough to startle and get specious debates going. This in turn will worsen the already acrimonious accusations about voting issues and election qualms.

The danger too is one of emboldening disenfranchisement.

If voters believe that AI has somehow aided in tainting or corrupting election activities (whether the belief is founded or unfounded), this alone could cause people to vote less and become disillusioned with voting all told. When people no longer believe that their vote is going to be properly counted, they usually correspondingly garner a sense of disbelief in the election results and democracy.

Franklin D. Roosevelt indicated the importance of awareness when it comes to our elections: “Democracy cannot succeed unless those who express their choice are prepared to choose wisely. The real safeguard of democracy, therefore, is education.”

AI will inch by inch find its way into the voting process. You should anticipate that this will be on a dual-use AI basis. We can seek to use AI as a bolstering force to ensure free and fair elections. Others will simultaneously be aiming to use AI as a means of corrupting our elections. It is essential that we become aware of and educated about how AI is going to enter into the voting process.

One last comment for now. I ardently vote that we use AI to aid and inspire free and fair elections while we remain alert to those that would wish instead to exploit AI for undercutting voting and our democracy.

I hope that everyone will vote with me on that inarguably vital proposition (except, of course, for the malefactors that perniciously desire otherwise).