Electronic Agents under the Law

1
Electronic Agents under the Law

• Reconciling the Promise of Agency Law Doctrine
  with Contract, Tort and Property Law Restrictions
  on Electronic Agent Deployment

2
Some Looming Issues

• Business practice divergence
• Who should utilize electronic agents
• Content providers vs. users
• Substantial transaction efficiency marketing
  opportunity enhancement captured by e-agents
• Use threatening to web content and service
  providers (disintermediation, excessive price
  visibility)
• However, restrictions on consumer use of
  electronic agents have substantial
  anti-competitive impact

3
Laws Applicable to eAgency

• Uniform Electronic Transactions Act (UETA)
• Electronic Signatures in Global and National
  Commerce Act (E-SIGN)
• Uniform Computer Information Transactions Act
  (UCITA)
• Other laws potentially governing use of eAgents
• Tort-tresspass, IP-infringement, Antitrust

4
Growing Significance of Electronic Agents

• Relatively new legal concept
• Legal recognition of electronic agents needed to
  legitimize emerging electronic commerce methods
  (manifestation of mutual assent)
• Common implementation (user interface to process
  consumer transactions on business-to-consumer
  (B2C) electronic commerce websites)
• Bots offer great potential for customer
  satisfaction economic efficiency, are often
  denied website access
• Seller implementation of electronic agents
• Buyer implementation of electronic agents

5
Bots vs. Search Engines

• Bots
• Bots serve as loyal agents for the benefit of the
  consumer user
• focus their efforts much more closely on
  satisfying only user criteria
• tendency towards personification - the essence of
  loyalty expected of a traditional
  human/human-directed agents
• Search Engines
• significant methodological flaws in the indexing,
  ranking and presentation of search results
• search results are also intentionally skewed to
  rank some results more highly than would be
  produced by objective ranking metrics
• Search engines perform much closer to the
  original, primitive conception of electronic
  agents as mere tools

6
Bots vs. Search Engines

• Key difference between bots and search engines
• Bots are more likely to produce results
  consistent with primarily addressing the consumer
  needs while search engines are more likely to
  produce results biased by their technical flaws
  other sources of conflicts of interest

7
Traditional vs. eAgency Law

• UETA envisions that the computer and information
  science communities will continue innovating to
  refine electronic agents
• Laws foresee electronic agents may evolve away
  from tools
• Roles such as negotiation, making judgments
  value assessments

8
Traditional vs. eAgency Law

• Eventually capable of implementing negotiation
  through the assessment of tradeoffs among various
  contract terms
• Computer science IST communities are pushing
  the envelope of electronic agents to employ
  judgment using expert systems and artificial
  intelligence

9
Traditional vs. eAgency Law

• Public Policy Context Electronic Agents
• UCITA UETA accept agency law
• Each envisions automated, mechanical or
  computerized communications
• Act as substitutes for human agents in
• Negotiation or conclusion of contracts
• Principal is bound by acts of eAgents in same
  situations as if human agents used

10
Electronic Agent under UETA

• Computer program, or electronic or other
  automated means
• Used by a person to initiate an action or
  respond to electronic messages or performances
• On the persons behalf
• Without review or action by an individual
• May manifest assent to a record or term after an
  opportunity for review

11
Automated Transactions

• Contracts formed without human intervention by
  one or both parties
• Assent may occur after eAgent starts performance
• Delivery of licensed information or
• Making a payment due
• Enabling review - communication must be in form
  that reasonably configured eAgent could be
  technically capable of using

12
Role of Humans vs. eAgents

• eAgents are semiautonomous
• Humans generally write configure underlying
  software
• Humans firms are principles benefited by eAgent
  activities
• eAgents may
• Perform preliminary negotiations
• Information delivery discovery
• Create contracts through mutual assent
• Offer, acceptance, document processing
• Performance of contracts
• Deliver electronic performances, incl Payments
• Document processing

13
Bots

• Shopping bots scour web info on goods given
  humans predetermined specifications
• Can site limit information discovery?
• Consumers approach perfect info
• Auction or shopping sites consider listings are
  proprietary data
• eBay v. Bidders Edge (N.D.Cal.2000)
• Bots trespass w/ continually requests for listing
  updates
• Do terms conditions of website use prevent
  automated comparison shopping?

14
(No Transcript)
15
Why are Bots Important?

• Bots Share Information Advantages Traditionally
  Held by Primary Aggregators with Secondary
  Aggregators for Shopping Benefit of Clients
• Bots Focus Search Engine Technology and Legal
  Rights on Databases
• Impact Rights to Post and Aggregate Content

16
Why are Bots Important?

• Emerging Legal Issues Span Various Fields
• Tort, Intellectual Property, License Assent
  Restrictions, Consumer Protection, Antitrust,
  eAgency
• Expect Significant Impact of Bots Regulation
  on
• Content Aggregation, P2P Technologies Markets,
  Monitoring Website Infringement, Automated
  Bargaining

17
Economic Impact of Bots

• Perfect Competition Model Efficient Markets
  Presume Perfect Information
• Bots Rely on Network Effects to Reduce Search
  Costs, Increase Comparisons, Approach Perfect
  Info
• Bots Facilitate Bidding Profits to Zero
• But, shoppers price-tunnel vision arguably
  ignores relevance of service
• Information Asymmetries Undermine Market
  Clearance, cause Adverse Selection

18
What are Sustainable Business Model(s) for Bots?

• Bot Innovation Unsustainable w/o Effective
  Revenue Model
• Innovation constantly needed to refine search,
  retrieval, analysis retain unfettered data
  access For Free then Fee once Audience is
  Captivated
• Online Infomediary or Personal Shopping Asst.
• Fee for Service
• Fixed/Variable per transaction, metered as
  what?!?
• Micro-Payments would permit initial growth
• Subscription Model
• Eyeballs Ad referrals enable mass
  personalization
• Slotting Preferences
• PII Profiling for Resale

19
Bots are Subset of Search Engine Architecture

• Bots have Many Operations Legal Issues Similar
  to Search Engines
• EX IP, Affiliation, Linking, Framing, Caching
• Shopping Bot a/k/a/ Crawler, Spider
• Autonomous Software Performs Search, Selection,
  Arrangement, Analysis, Linking
• Low Cost, Operates in Representative Role,
  Functionality Enabled by Network Effects

20
Disincentives to Tolerate Uncontrolled Bot
Activity

• Bots Could Bid Profits to Zero
• Historical Reluctance to Revealing Proprietary
  Information
• Game Theory Bargaining Strength Compromised
• Direct Comparisons Sales Lost to Competitors
• Eyeballs Diverted from Preferred Order of
  Impressions
• Bots Maliciously Invade Server Space
• Could be indistinguishable from hack/attacker

21
Three Models of Bot Interaction for Content
Sites

• Benign Tolerance
• Broaden exposure, increase traffic sales
• Cooperative
• Negotiate linking agreements Bot affiliation,
  slotting fees these approach form as
  distribution agreements
• Hostile Opposition
• Active technical legal maneuvers to prevent
  Bot probes, internalize value of postings,
  suppress non-competitive terms

22
Are Bots Cyber-Trespassers?

• Trespass to Chattels, Land or Sui Generis
• CompuServe v. Cyber Promotions, 962 F.Supp.1015
  (S.D.Oh.1997) (spam is trespass)
• Thrifty-Tel, Inc. v. Bezenek, 46 Cal.App.4th 1559
  (1996) (misappropriation of long distance
  service)
• eBay v. Bidders Edge, 100 F.Supp.2d 1058
  (N.D.Ca.2000) (PII, trespass, site congestion)
• Register.com v. Verio, 126 F.Supp.2d 238 (SDNY
  2000) (enjoined extraction of WHOIS info from
  dom. name registrar, congestion)
• Ticketmaster v. Tickets.com, (C.D.Cal.3.27.00)
• Computer Fraud Abuse Act, 18 USC 1030
• Intel v. Hamidi (Cal.S.Ct.2003) (no trespass to
  chattel unless email harms target computer
  server)

23
Do Bots Infringe IP?

• Kelly v. Arriba Soft, 2002 US App.LEXIS 1786 (9th
  Cir.2002) (Thumbnails - fair use)
• BT v. Prodigy, (UK patent on hyperlink)
• AM v. Napster, (contributory infringement,
  central database)
• NBA v. Motorola, 105 F.3d 841 (2d Cir. 1997)
  (state misappropriation narrowed)
• DMCA anti-circumvention
• Other Linking, Framing Caching Restraints
• Sui Generis Database Protections (e.g., EU)

24
Antitrust Consumer Protection

• Monopolization of Essential Facility
• Refusal to Deal bar info access to Bots
• Preventing Price Comparison is Anticompetitive,
  Erects Barriers to Entry
• Search Results Biased by Slotting Fees
• Unfair deceptive trade practice in obfuscating
  Bots bias
• Non-Objective Search Results Anti-Competitive

25
eAgents under Agency Law

• UETA, E-SIGN UCITA Encourage Equivalence of
  Electronic Agents
• Disclosure of Principles vs. Agents
• Anonymizer conceals principals identity
• Principals Rights/Needs to Use Agents
• Third partys Rights to Refuse to Deal w/
  Particular Agents or Classes of Agents

26
Robot Exclusion Standards

• Industry Convention (standard) as respect method
  to exclude search engines, bots spiders from
  copying or indexing content
• Defined Tutorals
• http//www.robotstxt.org/wc/norobots.html
• http//www.searchengineworld.com/robots/
• http//www.dwfaq.com/Tutorials/Miscellaneous/robot
  _txt.asp
• Googles index avoidance recommendation
• http//www.google.com/webmasters/remove.html

27
Enforceable Contracts Require Mutual Assent

• Mutual Assent Arises when
• Offer, acceptance, consideration, legal
  objective, S/F (writing vs. oral)
• Website Terms Conditions
• Licenses (EULA) are contracts too!
• Enforceability of shrink/click/browse wraps
• Nature of terms included (e.g., adhesion, waive
  fundamental rights)
• Robot exclusion?

28
e-Contracting Departures from Traditional
Practice

• Mutual Assent can be Implied from the parties
  conduct
• Implication strongest when regularly
  objectively discernable
• EX everybody expects to pay taxi fare
• Implication weaker in other contexts
• EX Tickets (cruise, sporting event), claim
  checks (pkg., coatroom), open source (Linux)

29
e-Contracting Departures from Traditional
Practice

• The 4 Wraps Shrink Click Box Browse
• Shrink-Wrap Software Agreements
• Accept standardized terms or return to decline
• Pro CD v. Zeidenberg, 86 F.3d 1447 (7th Cir.1996)
  (shrink wrap license use restriction enforceable)
  
• Increasingly enforceable because ubiquitous
• UCITA mass-market licenses
• Consumer must have opportunity to review
  archive terms

30
e-Contracting Departures from Traditional Practice

• Click-Wrap Agreements
• Software licenses appear when installed
• Specht v. Netscape
• Online license terms accessible before download
  installation of software or receipt of content
• eAgents work for content provider
• Terms in the Box
• Retaining hardware purchase implies acceptance of
  software license
• Gateway 2000

31
Browse Wrap/Click Free

• Mutual Assent implied from mere browsing
• Everyone surfer knows that there are terms
  conditions for website use
• BUT, stealth terms seldom equivalent to terms
  negotiated by market equals, risks unknown
• Cause surprise, unconscionable, adhesion,
  take-it-or-leave-it
• EX Robot Exclusion Standard?!?